Segui questo link per vedere altri tipi di pubblicazioni sul tema: Sentiment Analysis.
Tesi sul tema "Sentiment Analysis"
Cita una fonte nei formati APA, MLA, Chicago, Harvard e in molti altri stili
Vedi i top-50 saggi (tesi di laurea o di dottorato) per l'attività di ricerca sul tema "Sentiment Analysis".
Accanto a ogni fonte nell'elenco di riferimenti c'è un pulsante "Aggiungi alla bibliografia". Premilo e genereremo automaticamente la citazione bibliografica dell'opera scelta nello stile citazionale di cui hai bisogno: APA, MLA, Harvard, Chicago, Vancouver ecc.
Puoi anche scaricare il testo completo della pubblicazione scientifica nel formato .pdf e leggere online l'abstract (il sommario) dell'opera se è presente nei metadati.
Vedi le tesi di molte aree scientifiche e compila una bibliografia corretta.
1
Kavaliauskas, Vytautas. "Nuomonių analizės taikymas komentarams lietuvių kalboje". Master's thesis, Lithuanian Academic Libraries Network (LABT), 2011. http://vddb.laba.lt/obj/LT-eLABa-0001:E.02~2011~D_20110615_130252-94422.
Testo completoAbstract (sommario):
Pastaruosius keletą metų, žmonėms vis aktyviau pradėjus reikšti savo požiūrį, įsitikinimus ir potyrius internete, susiformavo nauja tyrinėjimų sritis, kuri apima nuomonių gavybą ir sentimentų analizę. Šios srities tyrinėjimus aktyviai skatina ir jais domisi įvairios verslo kompanijos, matančios didelį, dėka nuolat tobulėjančių rezultatų, praktinį potencialą. Šis darbas skirtas apžvelgti teorinius bei praktinius nuomonės gavybos ir sentimentų analizės rezultatus bei realizuoti prototipinę nuomonės analizės sistemą, skirtą tyrinėti trumpus komentarus, parašytus lietuvių kalba. Taip pat darbe aprašomos problemos, susijusios su lietuvių kalbos taikymu nuomonės gavybos ir sentimentų analizės sistemų veikloje. Galiausiai, baigiamojoje dalyje suformuluojami ir išdėstomi rekomendacinio pobūdžio etapai, skirti nuomonės analizės sistemų kūrimui bei tobulinimui.In past few years, more and more people started to express their views, beliefs and experiences on the Internet. This caused the emergence of a new research field, which includes opinion mining and sentiment analysis. Various business companies are actively interested in researches of this domain and seeing big potential for practical adaptation of the results. This Master Thesis covers the review of theoretical and practical results of opinion mining and sentiment analysis, including attempt of creating prototype system for opinion analysis of comments in Lithuanian. Also this study aims to identify problems related to adaptation of Lithuanian language in opinion mining and sentiment analysis system work. Finally, last part contains of the formulated guidance steps for development and improvement of the opinion mining and sentiment analysis.
2
Bao, Lingxian. "Sentiment induction for attention and lexicon regularized neural sentiment analysis: improving aspect-based neural sentiment analysis with lexicon enhancement, attention regularization and sentiment induction". Doctoral thesis, Universitat Pompeu Fabra, 2021. http://hdl.handle.net/10803/673363.
Testo completoAbstract (sommario):
Deep neural networks as an end-to-end approach lacks flexibility and robustness
from an application point of view, as one cannot easily adjust the
network to fix an obvious problem, especially when new training data is not
available: e.g. when the model consistently predicts positive when seeing
the word “disappointed”. Meanwhile, it is less stressed that the attention
mechanism is likely to “over-focus” on particular parts of a sentence, while
ignoring positions which provide key information for judging the polarity.
In this thesis, we describe a simple yet effective approach to leverage lexicon
information so that the model becomes more flexible and robust. We
also explore the effect of regularizing attention vectors to allow the network
to have a broader “focus” on the input sequence. Moreover, we try to
further improve the proposed lexicon enhanced neural sentiment analysis
system by applying sentiment domain adaptation.Las redes neuronales profundas como enfoque integral carecen de flexibilidad y robustez desde el punto de vista de la aplicación, ya que no se puede ajustar fácilmente la red para solucionar un problema evidente, especialmente cuando no se dispone de nuevos datos de entrenamiento: por ejemplo, cuando el modelo predice sistemáticamente positivo al ver la palabra ”decepcionado”. Por otro lado, se hace menos hincapié en que es probable que el mecanismo de atención "se concentre demasiado” en partes concretas de una oración, mientras ignora posiciones que proporcionan información clave para juzgar la polaridad. En esta tesis, describimos un enfoque sencillo pero eficaz para aprovechar la información del léxico de modo que el modelo sea maás flexible y robusto. También exploramos el efecto de regularizar los vectores de atención para permitir que la red tenga un ”enfoque” más amplio en la secuencia de entrada. Además, tratamos de mejorar aún más el sistema que proponemos de análisis profundo de sentimiento con el soporte de léxico aplicando sobre el mismo la adaptación del anàlisis de sentimiento al dominio.
深度神经网络作为一种端到端的方法，从应用的角度来看缺乏灵 活性和鲁棒性。例如，当模型看到词语“失望”却始终预测正值时， 在没有新的训练数据的情况下，很难轻易地通过调整模型来解决问 题。另外，常用的注意力机制可能会“过度关注”句子的某些特定部 分，从而忽略能提供判断极性关键信息的位置；此情况在业界鲜有 提及。在本文中，我们描述一种简单却行之有效的方法将词典信息 与深度神经网络相结合，从而改进模型的灵活性及鲁棒性。我们亦 探索通过正则化注意力向量来抑制注意力机制“过度关注”的问题。 此外，我们尝试通过应用情感域自适应来进一步改进所提出的词典 增强型神经情感分析系统。
3
Erogul, Umut. "Sentiment Analysis In Turkish". Master's thesis, METU, 2009. http://etd.lib.metu.edu.tr/upload/12610616/index.pdf.
Testo completoAbstract (sommario):
Sentiment analysis is the automatic classification of a text, trying to determine the attitude of the writer with respect to a specific topic. The attitude may be either their judgment or evaluation, their feelings or the intended emotional communication.
The recent increase in the use of review sites and blogs, has made a great amount of subjective data available. Nowadays, it is nearly impossible to manually process all the relevant data available, and as a consequence, the importance given to the automatic classification of unformatted data, has increased.
Up to date, all of the research carried on sentiment analysis was focused on English language. In this thesis, two Turkish datasets tagged with sentiment information is introduced and existing methods for English are applied on these datasets. This thesis also suggests new methods for Turkish sentiment analysis.
4
Cheng, Tai Wai David. "Corpus and sentiment analysis". Thesis, University of Surrey, 2007. http://epubs.surrey.ac.uk/2744/.
Testo completoAbstract (sommario):
Information extraction/retrieval has been of interest to researchers since the early 1960's. A series of conferences and competitions have been held by DARPA/NIST since the late 1980's has resulted in the analysis of news reports and government reports in English and other languages, notably Chinese and Arabic. A number of methods have been developed for analysing `free' natural language texts. Furthermore, a number of systems for understanding messages have been developed, focusing on the area of named entity extraction, templates for dealing with certain kinds of news. The templates were handcrafted, and a lot of ad-hoc knowledge went into the creation of such systems. Seven of these systems have been reviewed. Despite the fact that IE systems built for different tasks often differ from each other, the core elements are shared by nearly every extraction system. Some of these core elements such as parser and part of speech (POS) tagger, are tuned for optimal performance for a specific domain, or text with pre-defined structures. The extensive use of gazetteers and manually crafted grammar rules further limits the portability of the existing IE systems to work language and domain independently. The goal of this thesis is to develop an algorithm that can be used to extract information from free texts, in our case, from financial news text; and from arbitrary domains unambiguously. We believe the use of corpus linguistics and statistical techniques would be more appropriate and efficient for this task, instead of using other approaches that rely on machine learning, POS taggers, parsers, and so on, which are tuned to work for a predefined domain. Based on this belief, a framework using corpus linguistics and statistical techniques, to extract information as unambiguously as possible from arbitrary domains was developed. A contrastive evaluation has been carried out not only in the domain of financial texts and movie reviews, but also with multi-lingual texts (Chinese and English). The results are encouraging. Our preliminarily evaluation, based on the correlation between a time series of positive (negative) sentiment word (phrase) counts with a time series of indices produced by stock exchanges (Financial Times Stock Exchange, Dow Jones Industrial Average, Nasdaq, S&P 500, Hang Seng Index, Shanghai Index, and Shenzhen Index) showed that when the positive (negative) sentiment series correlates with the stock exchange index, the negative (positive) shows a smaller degree of correlation and in many cases a degree of anti-correlation. Any interpretation of our result requires a careful econometrically well grounded analysis of the financial time series - this is beyond the scope of this work.
5
Niccum, Cameron Michael. "Sentiment Analysis using Tensor2Tensor". OpenSIUC, 2018. https://opensiuc.lib.siu.edu/theses/2440.
Testo completoAbstract (sommario):
Tensor2Tensor is a library of deep learning models and datasets developed by the Google Brain team. The transformer models of this library are implemented using stacked self-attention layers instead of containing recurrent layers. Two of these transformer models have been chosen along with a few hyperparameter sets for analyzing the Internet Movie Database (IMDB). The runtime and accuracy of these models are then compared to themselves as well as the accuracy of previously known models.
6
Melloncelli, Damiano. "Sentiment analysis in Twitter". Master's thesis, Alma Mater Studiorum - Università di Bologna, 2014. http://amslaurea.unibo.it/6592/.
Testo completoAbstract (sommario):
Gli ultimi anni hanno visto una crescita esponenziale nell’uso dei social media (recensioni, forum, discussioni, blog e social network); le persone e le aziende utilizzano sempre più le informazioni (opinioni e preferenze) pubblicate in questi mezzi per il loro processo decisionale. Tuttavia, il monitoraggio e la ricerca di opinioni sul Web da parte di un utente o azienda risulta essere un problema molto arduo a causa della proliferazione di migliaia di siti; in più ogni sito contiene un enorme volume di testo non sempre decifrabile in maniera ottimale (pensiamo ai lunghi messaggi di forum e blog). Inoltre, è anche noto che l’analisi soggettiva delle informazioni testuali è passibile di notevoli distorsioni, ad esempio, le persone tendono a prestare maggiore attenzione e interesse alle opinioni che risultano coerenti alle proprie attitudini e preferenze. Risulta quindi necessario l’utilizzo di sistemi automatizzati di Opinion Mining, per superare pregiudizi soggettivi e limitazioni mentali, al fine di giungere ad una metodologia di Sentiment Analysis il più possibile oggettiva.
7
Filho, Pedro Paulo Balage. "Aspect extraction in sentiment analysis for portuguese language". Universidade de São Paulo, 2017. http://www.teses.usp.br/teses/disponiveis/55/55134/tde-05122017-140435/.
Testo completoAbstract (sommario):
Aspect-based sentiment analysis is the field of study which extracts and interpret the sentiment, usually classified as positive or negative, towards some target or aspect in an opinionated text. This doctoral dissertation details an empirical study of techniques and methods for aspect extraction in aspect-based sentiment analysis with the focus on Portuguese. Three different approaches were explored: frequency-based, relation-based and machine learning. In each one, this work shows a comparative study between a Portuguese and an English corpora and the differences found in applying the approaches. In addition, richer linguistic knowledge is also explored by using syntatic dependencies and semantic roles, leading to better results. This work lead to the establishment of new benchmarks for the aspect extraction in Portuguese.A análise do sentimento orientada a aspectos é o campo de estudo que extrai e interpreta o sentimento, geralmente classificado como positivo ou negativo, em direção a algum alvo ou aspecto em um texto de opinião. Esta tese de doutorado detalha um estudo empírico de técnicas e métodos para extração de aspectos em análises de sentimentos baseadas em aspectos com foco na língua Portuguesa. Foram exploradas três diferentes abordagens: métodos baseados na frequências, métodos baseados na relação e métodos de aprendizagem de máquina. Em cada abordagem, este trabalho mostra um estudo comparativo entre um córpus para o Português e outro para o Inglês e as diferenças encontradas na aplicação destas abordagens. Além disso, o conhecimento linguístico mais rico também é explorado pelo uso de dependências sintáticas e papéis semânticos, levando a melhores resultados. Este trabalho resultou no estabelecimento de novos padrões de avaliação para a extração de aspectos em Português.
8
Gaudette, Lisa. "Compact features for sentiment analysis". Thesis, University of Ottawa (Canada), 2009. http://hdl.handle.net/10393/28295.
Testo completoAbstract (sommario):
This work examines a novel method of developing features to use for machine learning of sentiment analysis and related tasks. This task is frequently approached using a Bag of Words representation -- one feature for each word encountered in the training data -- which can easily number in the thousands or tens of thousands. This thesis develops a set of "numeric" features, by learning scores for words, dividing the range of possible scores into a number of bins, and then generating features based on counting how many words in each document have scores in each bin. This allows for effective learning of sentiment and related tasks with 25 features; in fact, performance was very often slightly better with these features. This reduction in the number of features allows for the processing of much larger collections of texts than previously attempted. In addition, we carefully consider the problem of evaluating ordinal problems.
9
Athar, Awais. "Sentiment analysis of scientific citations". Thesis, University of Cambridge, 2014. https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.707942.
Testo completo
10
Smith, Phillip. "Sentiment analysis of patient feedback". Thesis, University of Birmingham, 2017. http://etheses.bham.ac.uk//id/eprint/7406/.
Testo completoAbstract (sommario):
The application of sentiment analysis as a method for the automatic categorisation of opinions in text has grown increasingly popular across a number of domains over the past few years. In particular, health services have started to consider sentiment analysis as a solution for the task of processing the ever-growing amount of feedback that is received in regards to patient care. However, the domain is relatively under-studied in regards to the application of the technology, and the effectiveness and performance of methods have not been substantially demonstrated. Beginning with a survey of sentiment analysis and an examination of the work undertaken so far in the clinical domain, this thesis examines the application of supervised machine learning models to the classification of sentiment in patient feedback. As a starting point, this requires a suitably annotated patient feedback dataset, for both analysis and experimentation. Following the construction and detailed analysis of such a resource, a series of machine learning experiments study the impact of different models, features and review types to the problem. These experiments examine the applicability of the selected methods and demonstrate that model and feature choice may not be a significant issue in sentiment classification, whereas the type of review that the models train and test across does affect the outcome of classification. Finally, by examining the role that responses play in the patient feedback process and developing the idea of incorporating the inter-document context provided by the response into the feedback classification process, a recalibration framework for the labelling of sentiment in ambiguous texts for patient feedback is developed. As this detailed analysis will demonstrate, while some problems in performance remain despite the development and implementation of the recalibration framework, sentiment analysis of patient feedback is indeed viable, and achieves a classification accuracy of 91.4% and F1 of 0.902 on the gathered data. Furthermore, the models and data can serve as a baseline to study the nature of patient feedback, and provide a unique opportunity for the development of sentiment analysis in the clinical domain.
11
Hamdan, Hussam. "Sentiment analysis in social media". Thesis, Aix-Marseille, 2015. http://www.theses.fr/2015AIXM4356.
Testo completoAbstract (sommario):
Dans cette thèse, nous abordons le problème de l'analyse des sentiments. Plus précisément, nous sommes intéressés à analyser le sentiment exprimé dans les textes de médias sociaux.Nous allons nous concentrer sur deux tâches principales: la détection de polarité de sentiment dans laquelle nous cherchons à déterminer la polarité (positive, négative ou neutre) d'un texte donné et l'extraction de cibles d’opinion et le sentiment exprimé vers ces cibles (par exemple, pour le restaurant nous allons extraire des cibles comme la nourriture, pizza, service). Notre principal objectif est de construire des systèmes à la pointe de la technologie qui pourrait faire les deux tâches. Par conséquent, nous avons proposé des systèmes supervisés différents suivants trois axes de recherche: l'amélioration de la performance du système par la pondération de termes, en enrichissant de la représentation de documents et en proposant un nouveau modèle pour la classification de sentiment.Pour l'évaluation, nous avons participé à un atelier international sur l'évaluation sémantique (Sem Eval), nous avons choisi deux tâches: l'analyse du sentiment sur Twitter dans laquelle nous déterminer la polarité d'un tweet et l'analyse des sentiments basée sur l’aspect dans laquelle nous extrayons les cibles d'opinion dans les critiques de restaurants, puis nous déterminons la polarité de chaque cible, nos systèmes ont été classés parmi les premiers trois meilleurs systèmes dans toutes les sous-tâches. Nous avons également appliqué nos systèmes sur un corpus des critiques de livres français construit par l'équipe Open Edition pour extraire les cibles d'opinion et leurs polaritésIn this thesis, we address the problem of sentiment analysis. More specifically, we are interested in analyzing the sentiment expressed in social media texts such as tweets or customer reviews about restaurant, laptop, hotel or the scholarly book reviews written by experts. We focus on two main tasks: sentiment polarity detection in which we aim to determine the polarity (positive, negative or neutral) of a given text and the opinion target extraction in which we aim to extract the targets that the people tend to express their opinions towards them (e.g. for restaurant we may extract targets as food, pizza, service).Our main objective is constructing state-of-the-art systems which could do the two tasks. Therefore, we have proposed different supervised systems following three research directions: improving the system performance by term weighting, by enriching the document representation and by proposing a new model for sentiment classification. For evaluation purpose, we have participated at an International Workshop on Semantic Evaluation (SemEval), we have chosen two tasks: Sentiment analysis in twitter in which we determine the polarity of a tweet and Aspect-Based sentiment analysis in which we extract the opinion targets in restaurant reviews, then we determine the polarity of each target. Our systems have been among the first three best systems in all subtasks. We also applied our systems on a French book reviews corpus constructed by OpenEdition team for extracting the opinion targets and their polarities
12
Moilanen, Karo. "Compositional entity-level sentiment analysis". Thesis, University of Oxford, 2010. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.559817.
Testo completoAbstract (sommario):
This thesis presents a computational text analysis tool called AFFECTiS (Affect Interpretation/Inference System) which focuses on the task of interpreting natural language text based on its subjective, non-factual, affective properties that go beyond the 'traditional' factual, objective dimensions of meaning that have so far been the main focus of Natural Language Processing and Computational Linguistics. The thesis presents a fully compositional uniform wide-coverage computational model of sentiment in text that builds on a number of fundamental compositional sentiment phenomena and processes discovered by detailed linguistic analysis of the behaviour of sentiment across key syntactic constructions in English. Driven by the Principle of Semantic Compositionality, the proposed model breaks sentiment interpretation down into strictly binary combinatory steps each of which explains the polarity of a given sentiment expression as a function of the properties of the sentiment carriers contained in it and the grammatical and semantic context(s) involved. An initial implementation of the proposed compositional sentiment model is de- scribed which attempts direct logical sentiment reasoning rather than basing compu- tational sentiment judgements on indirect data-driven evidence. Together with deep grammatical analysis and large hand-written sentiment lexica, the model is applied recursively to assign sentiment to all (sub )sentential structural constituents and to concurrently equip all individual entity mentions with gradient sentiment scores. The system was evaluated on an extensive multi-level and multi-task evaluation framework encompassing over 119,000 test cases from which detailed empirical ex- perimental evidence is drawn. The results across entity-, phrase-, sentence-, word-, and document-level data sets demonstrate that AFFECTiS is capable of human-like sentiment reasoning and can interpret sentiment in a way that is not only coherent syntactically but also defensible logically - even in the presence of the many am- biguous extralinguistic, paralogical, and mixed sentiment anomalies that so tellingly characterise the challenges involved in non-factual classification.
13
Saif, Hassan. "Semantic sentiment analysis of microblogs". Thesis, Open University, 2015. http://oro.open.ac.uk/44063/.
Testo completoAbstract (sommario):
Microblogs and social media platforms are now considered among the most popular forms of online communication. Through a platform like Twitter, much information reflecting people's opinions and attitudes is published and shared among users on a daily basis. This has recently brought great opportunities to companies interested in tracking and monitoring the reputation of their brands and businesses, and to policy makers and politicians to support their assessment of public opinions about their policies or political issues. A wide range of approaches to sentiment analysis on Twitter, and other similar microblogging platforms, have been recently built. Most of these approaches rely mainly on the presence of affect words or syntactic structures that explicitly and unambiguously reflect sentiment (e.g., "great'', "terrible''). However, these approaches are semantically weak, that is, they do not account for the semantics of words when detecting their sentiment in text. This is problematic since the sentiment of words, in many cases, is associated with their semantics, either along the context they occur within (e.g., "great'' is negative in the context "pain'') or the conceptual meaning associated with the words (e.g., "Ebola" is negative when its associated semantic concept is "Virus"). This thesis investigates the role of words' semantics in sentiment analysis of microblogs, aiming mainly at addressing the above problem. In particular, Twitter is used as a case study of microblogging platforms to investigate whether capturing the sentiment of words with respect to their semantics leads to more accurate sentiment analysis models on Twitter. To this end, several approaches are proposed in this thesis for extracting and incorporating two types of word semantics for sentiment analysis: contextual semantics (i.e., semantics captured from words' co-occurrences) and conceptual semantics (i.e., semantics extracted from external knowledge sources). Experiments are conducted with both types of semantics by assessing their impact in three popular sentiment analysis tasks on Twitter; entity-level sentiment analysis, tweet-level sentiment analysis and context-sensitive sentiment lexicon adaptation. Evaluation under each sentiment analysis task includes several sentiment lexicons, and up to 9 Twitter datasets of different characteristics, as well as comparing against several state-of-the-art sentiment analysis approaches widely used in the literature. The findings from this body of work demonstrate the value of using semantics in sentiment analysis on Twitter. The proposed approaches, which consider words' semantics for sentiment analysis at both, entity and tweet levels, surpass non-semantic approaches in most datasets.
14
Liu, Qiaoshan. "Sentiment analysis in social events". Thesis, Linnéuniversitetet, Institutionen för samhällsstudier (SS), 2018. http://urn.kb.se/resolve?urn=urn:nbn:se:lnu:diva-78077.
Testo completoAbstract (sommario):
Purpose: The purpose of this study is going to visualize the public sentiment on expected and unexpected social events. Exploring the relationship between tweets forwarding and sentiment. Design/methodology/approach: This research related to sentiment analysis of social events applied a lexicon-based method. The social events come from Facebook data breach and Ireland vote on abortion event. The study conducted This study focused on how the public sentiment changes over time and the relationship between sentiment and tweet forwarding. Bing lexicon and NRC lexicon are adopted in the analysis. Result: The result of this study is the dominant sentiment trend is consistent with the trend of the number of tweets over time in the Facebook data breach and Ireland vote on abortion. Besides, the sentiment has affected people forward tweets in this research.
15
Pessato, Luca <1990>. "Social Media e Sentiment Analysis". Master's Degree Thesis, Università Ca' Foscari Venezia, 2016. http://hdl.handle.net/10579/9137.
Testo completoAbstract (sommario):
La tesi cerca di spiegare in che modo le nuove tecnologie e gli strumenti del web riescano ad influenzare i comportamenti e le opinioni di chi li utilizza, dagli utenti alle aziende di tutto il mondo. Si introdurranno i nuovi metodi per analizzare e capire il sentiment generato da ciò che viene scritto e pubblicato sui principali social media.
Si definiranno inizialmente i concetti di internet e rivoluzione web 2.0, spiegando quali cambiamenti essi hanno portato nella società e qual’ è stata l’evoluzione e la crescita nel tempo.
Successivamente si descriveranno i Social Media, concentrando l’attenzione sulla loro diffusione globale e su quali piattaforme web assumano più importanza all’interno di questo insieme. Si elencheranno quelli di maggior impatto e successo tra gli utenti e le aziende, esplicitandone le differenti caratteristiche e i metodi di utilizzo.
In seguito si analizzeranno i maggiori cambiamenti in ambito strategico, che le imprese hanno adottato con il boom del web e dei social, con riferimento al nuovo modo di interazione tra produttore e consumatore, e su quale sia l’approccio migliore per avere successo on line.
Il Social Media Mining e la Sentimet Analysis saranno spiegati in dettaglio, concentrando l’attenzione su come essi possano essere sfruttati per ottenere conoscenza da dati grezzi e da testi scritti, contenenti informazioni importanti sui comportamenti delle persone e sulle loro opinioni spontanee. Si esporranno i procedimenti di queste recenti metodologie, che servono per analizzare, sia quantitativamente che qualitativamente, come le persone si comportano e cosa pensano di un determinato argomento, prodotto o servizio.
Si capirà, infine, come le azioni degli utenti e le informazioni che si possono trovare in internet abbiano assunto e assumeranno in futuro una notevole importanza per l’analisi dei dati e le imprese, poiché essi rappresentano la vera ricchezza del web.
16
YADAV, DEEPIKA. "SENTIMENT ANALYSIS ON TWITTER DATA". Thesis, DELHI TECHNOLOGICAL UNIVERSITY, 2020. http://dspace.dtu.ac.in:8080/jspui/handle/repository/18821.
Testo completoAbstract (sommario):
Prior to purchasing an item, individuals for the most part go to different shops in the market, question about the item, cost, and guarantee, and afterward at long last purchase the item dependent on the feelings they got on cost and nature of administration. This procedure is tedious and the odds of being cheated by the merchant are more as there is no one to direct regarding where the purchaser can get valid item and with legitimate expense. Be that as it may, presently a-days a decent number of people rely upon the upon line showcase for purchasing their necessary items. This is on the grounds that the data about the items is accessible from numerous sources; in this manner, it is relatively modest and furthermore has the office of home conveyance. Once more, before experiencing the way toward setting request for any item, clients all the time allude to the remarks or audits of the current clients of the item, which assist them with taking choice about the nature of the item just as the administration gave by the dealer. Like putting request for items, it is seen that there are many experts in the field of films, who experience the film and afterward at long last give a remark about the nature of the film, i.e., to watch the film or not or in five-star rating. These audits are basically in the content arrangement and at times extreme to comprehend. In this manner, these reports should be prepared suitably to get some important data. Order of these audits is one of the ways to deal with extricate information about the surveys. In this theory, distinctive AI procedures are utilized to characterize the audits. Reproduction and trials are done to assess the exhibition of the proposed grouping strategies.
It is seen that a decent number of scientists have frequently thought to be two distinctive survey datasets for conclusion grouping to be specific ascension and Polarity dataset. The IMDb dataset is separated into preparing and testing information. Accordingly, preparing information are utilized for preparing the AI calculations and testing information are utilized to test the information dependent on the preparation data. Then again, extremity dataset doesn't have separate information for preparing and testing. In this way, k-crease cross approval procedure is utilized to order the surveys. Four diverse AI strategies (MLTs) viz., Naive Bayes (NB), Support Vector Machine (SVM), Random Forest (RF), and Linear Discriminant Analysis (LDA) are utilized for the order of these film audits. Diverse execution assessment boundaries are utilized to assess the presentation of the AI strategies. It is seen that among the over four AI calculations, RF method yields the grouping result, with more precision.
Also, n-gram based characterization of surveys is completed on the ascension dataset.
v
The distinctive n-gram procedures utilized are unigram, bigram, trigram, unigram bigram, bigram + trigram, unigram + bigram + trigram. Four distinctive AI strategies, for example, Naive Bayes (NB), Maximum Entropy (ME), Support Vector Machine (SVM), and Stochastic Gradient Descent (SGD) methods are utilized to arrange the film surveys dependent on the n-gram strategy as referenced before. Diverse execution assessment boundaries are utilized to assess the presentation of these AI methods. The SVM method with unigram + bigram approach has demonstrated more exact outcome among every different methodologies.
Thirdly, SVM-based element determination strategy is utilized to choose best highlights from the arrangement everything being equal. These chose highlights are then considered as contribution to Artificial Neural Network (ANN) to characterize the surveys information. For this situation, two distinctive audit datasets i.e., IMDb and Polarity dataset are considered for grouping. In this technique, each expression of these surveys is considered as a component, and the assumption estimation of each word is determined. The component choice is done dependent on the opinion estimations of the expression. The words having higher assumption esteems are chosen. These words at that point go about as a contribution to ANN based on which the film audits are ordered.
At last, Genetic Algorithm (GA) is utilized to speak to the film surveys as chromosomes. Various activities of GA are completed to get the last arrangement result. Alongside this, the GA is likewise utilized as highlight choice to choose the best highlights from the arrangement of all highlights which in the end are given as contribution to ANN to acquire the last grouping outcome. Distinctive execution assessment boundaries are utilized to assess the presentation of GA and half breed of GA with ANN.
Feeling examination regularly manages investigation of surveys, remarks about any item, which are for the most part printed in nature and need legitimate preparing to got any significant data. In this postulation, various methodologies have been proposed to arrange the audits into particular extremity gatherings, i.e., positive and negative. Distinctive MLTs are utilized in this theory to play out the errand of arrangement and execution of every strategy is assessed by utilizing various boundaries, viz., exactness, review, f-measure and precision. The outcomes acquired by the proposed approaches are seen as better than the outcomes as announced by different creators in writing utilizing same dataset and approaches.
17
Svensson, Kristoffer. "Sentiment Analysis With Convolutional Neural Networks : Classifying sentiment in Swedish reviews". Thesis, Linnéuniversitetet, Institutionen för datavetenskap (DV), 2017. http://urn.kb.se/resolve?urn=urn:nbn:se:lnu:diva-64768.
Testo completoAbstract (sommario):
Today many companies exist and market their products and services on social medias, and therefore may receive reviews and thoughts from their end-users directly in these social medias. Reading every text by hand can be time-consuming, so by analysing the sentiment for all texts give the companies an overview how positive or negative the users are on a specific subject. Sentiment analysis is a feature that Beanloop AB is interested in implementing in their future projects and this thesis research problem was to investigate how deep learning could be used for this task. It was done by conducting an experiment with deep learning and neural networks. Several convolutional neural network models were implemented with different settings to find a combination of settings that gave the highest accuracy on the given test dataset. There were two different kind of models, one kind classifying positive and negative, and the second classified the previous two categories but also neutral. The training dataset and the test dataset contained data from two recommendation sites, www.reco.se and se.trustpilot.com. The final result shows that when classifying three categories (positive, negative and neutral) the models had problems to reach an accuracy at 85%, were only one model reached 80% accuracy as best on the test dataset. However, when only classifying two categories (positive and negative) the models showed very good results and reached almost 95% accuracy for every model.
18
Silva, Nadia Felix Felipe da. "Análise de sentimentos em textos curtos provenientes de redes sociais". Universidade de São Paulo, 2016. http://www.teses.usp.br/teses/disponiveis/55/55134/tde-27092016-143947/.
Testo completoAbstract (sommario):
A análise de sentimentos é um campo de estudo com recente popularização devido ao crescimento da Internet e do conteúdo que é gerado por seus usuários, principalmente nas redes sociais, nas quais as pessoas publicam suas opiniões em uma linguagem coloquial e em muitos casos utilizando de artifícios gráficos para tornar ainda mais sucintos seus diálogos. Esse cenário é observado no Twitter, uma ferramenta de comunicação que pode facilmente ser usada como fonte de informação para várias ferramentas automáticas de inferência de sentimentos. Esforços de pesquisas têm sido direcionados para tratar o problema de análise de sentimentos em redes sociais sob o ponto de vista de um problema de classificação, com pouco consenso sobre qual é o classificador com melhor poder preditivo, bem como qual é a configuração fornecida pela engenharia de atributos que melhor representa os textos. Outro problema é que em um cenário supervisionado, para a etapa de treinamento do modelo de classificação, é imprescindível se dispor de exemplos rotulados, uma tarefa árdua e que demanda esforço humano em grande parte das aplicações. Esta tese tem por objetivo investigar o uso de agregadores de classificadores (classifier ensembles), explorando a diversidade e a potencialidade de várias abordagens supervisionadas quando estas atuam em conjunto, além de um estudo detalhado da fase que antecede a escolha do classificador, a qual é conhecida como engenharia de atributos. Além destes aspectos, um estudo mostrando que o aprendizado não supervisionado pode fornecer restrições complementares úteis para melhorar a capacidade de generalização de classificadores de sentimento é realizado, fornecendo evidências de que ganhos já observados em outras áreas do conhecimento também podem ser obtidos no domínio em questão. A partir dos promissores resultados experimentais obtidos no cenário de aprendizado supervisionado, alavancados pelo uso de técnicas não supervisionadas, um algoritmo existente, denominado de C3E (Consensus between Classification and Clustering Ensembles) foi adaptado e estendido para o cenário semissupervisionado. Este algoritmo refina a classificação de sentimentos a partir de informações adicionais providas pelo agrupamento em um procedimento de autotreinamento (self-training). Tal abordagem apresenta resultados promissores e competitivos com abordagens que representam o estado da arte em outros domínios.Sentiment analysis is a field of study that shows recent popularization due to the growth of Internet and the content that is generated by its users. More recently, social networks have emerged, where people post their opinions in colloquial and compact language. This is what happens in Twitter, a communication tool that can easily be used as a source of information for various automatic tools of sentiment inference. Research efforts have been directed to deal with the problem of sentiment analysis in social networks from the point of view of a classification problem, where there is no consensus about what is the best classifier, and what is the best configuration provided by the feature engineering process. Another problem is that in a supervised setting, for the training stage of the classification model, we need labeled examples, which are hard to get in the most of applications. The objective of this thesis is to investigate the use of classifier ensembles, exploring the diversity and the potential of various supervised approaches when these work together, as well as to provide a study about the phase that precedes the choice of the classifier, which is known as feature engineering. In addition to these aspects, a study showing that unsupervised learning techniques can provide useful and additional constraints to improve the ability of generalization of the classifiers is also carried out. Based on the promising results got in supervised learning settings, an existing algorithm called C3E (Consensus between Classification and Clustering Ensembles) was adapted and extended for the semi-supervised setting. This algorithm refines the sentiment classification from additional information provided by clusters of data, in a self-training procedure. This approach shows promising results when compared with state of the art algorithms.
19
Vargas, Danny Suarez. "Detecting contrastive sentences for sentiment analysis". reponame:Biblioteca Digital de Teses e Dissertações da UFRGS, 2016. http://hdl.handle.net/10183/148304.
Testo completoAbstract (sommario):
A análise de contradições é uma área relativamente nova, multidisciplinar e complexa que tem por objetivo principal identificar pedaços contraditórios de texto. Ela pode ser abordada a partir das perspectivas de diferentes áreas de pesquisa, tais como processamento de linguagem natural, mineração de opinioes, recuperação de informações e extração de Informações. Este trabalho foca no problema de detectar contradições em textos – mais especificamente, nas contradições que são o resultado da diversidade de sentimentos entre as sentenças de um determinado texto. Ao contrário de outros tipos de contradições, a detecção de contradições baseada em sentimentos pode ser abordada como uma etapa de pós-processamento na tarefa tradicional de análise de sentimentos. Neste contexto, este trabalho apresenta duas contribuições principais. A primeira é um estudo exploratório da tarefa de classificação, na qual identificamos e usamos diferentes ferramentas e recursos. A segunda contribuição é a adaptação e a extensão de um framework de análise contradição existente, filtrando seus resultados para remover os comentários erroneamente rotulados como contraditórios. O método de filtragem baseia-se em dois algoritmos simples de similaridade entre palavras. Uma avaliação experimental em comentários sobre produtos reais mostrou melhorias proporcionais de até 30 % na acurácia da classificação e 26 % na precisão da detecção de contradições.Contradiction Analysis is a relatively new multidisciplinary and complex area with the main goal of identifying contradictory pieces of text. It can be addressed from the perspectives of different research areas such as Natural Language Processing, Opinion Mining, Information Retrieval, and Information Extraction. This work focuses on the problem of detecting sentiment-based contradictions which occur in the sentences of a given review text. Unlike other types of contradictions, the detection of sentiment-based contradictions can be tackled as a post-processing step in the traditional sentiment analysis task. In this context, we make two main contributions. The first is an exploratory study of the classification task, in which we identify and use different tools and resources. Our second contribution is adapting and extending an existing contradiction analysis framework by filtering its results to remove the reviews that are erroneously labeled as contradictory. The filtering method is based on two simple term similarity algorithms. An experimental evaluation on real product reviews has shown proportional improvements of up to 30% in classification accuracy and 26% in the precision of contradiction detection.
20
Pak, Alexander. "Automatic, adaptive, and applicative sentiment analysis". Phd thesis, Université Paris Sud - Paris XI, 2012. http://tel.archives-ouvertes.fr/tel-00717329.
Testo completoAbstract (sommario):
Sentiment analysis is a challenging task today for computational linguistics. Because of the rise of the social Web, both the research and the industry are interested in automatic processing of opinions in text. In this work, we assume a multilingual and multidomain environment and aim at automatic and adaptive polarity classification.We propose a method for automatic construction of multilingual affective lexicons from microblogging to cover the lack of lexical resources. To test our method, we have collected over 2 million messages from Twitter, the largest microblogging platform, and have constructed affective resources in English, French, Spanish, and Chinese.We propose a text representation model based on dependency parse trees to replace a traditional n-grams model. In our model, we use dependency triples to form n-gram like features. We believe this representation covers the loss of information when assuming independence of words in the bag-of-words approach.Finally, we investigate the impact of entity-specific features on classification of minor opinions and propose normalization schemes for improving polarity classification. The proposed normalization schemes gives more weight to terms expressing sentiments and lower the importance of noisy features.The effectiveness of our approach has been proved in experimental evaluations that we have performed across multiple domains (movies, product reviews, news, blog posts) and multiple languages (English, French, Russian, Spanish, Chinese) including official participation in several international evaluation campaigns (SemEval'10, ROMIP'11, I2B2'11).
21
Dubois, Damien. "Sentiment analysis: Transferring knowledge across domains". Thesis, Fredericton: University of New Brunswick, 2012. http://hdl.handle.net/1882/44592.
Testo completoAbstract (sommario):
This research focuses on text data mining and more precisely sentiment anal ysis algorithms. Sentiment analysis is a method of extracting sentiment knowledge out of unstructured text documents. This task raises many important challenges among which is domain dependency. Sentiment classi cation algorithms trained on one domain do not perform as well in another domain. The purpose of this thesis is to compare solutions of the well known challenges in sentiment analysis and to propose a solution to deal with the domain dependency issue. To do this, the framework developed in this thesis is intended to transfer the knowledge from a speci c domain or general knowledge to another speci c domain using the Min-cut algorithm. This algorithm can be used to mix the knowledge extracted using machine learning or general methods and that extracted using domain dependent methods.
The goal is then to improve the accuracy obtained with usual sentiment analysis algorithms on each domain. The proposed approach will be implemented in a general system that will have as input a corpus of unlabeled documents
dealing with a speci c topic such as movies and will return as output a grade of positivity for each document. In order to compare the solution for other challenges such as feature selection or negation, the framework will also take as input the method to deal with these challenges.
22
Serrano, Melissa. "Bilingual Sentiment Analysis of Spanglish Tweets". Thesis, Florida Atlantic University, 2017. http://pqdtopen.proquest.com/#viewpdf?dispub=10610508.
Testo completoAbstract (sommario):
Sentiment Analysis has been researched in a variety of contexts but in this thesis, the focus is on sentiment analysis in Twitter, which poses its own unique challenges such as the use of slang, abbreviations, emoticons, hashtags, and user mentions. The 140-character restriction on the length of tweets can also lead to text that is difficult even for a human to determine its sentiment. Specifically, this study will analyze sentiment analysis of bilingual (U.S. English and Spanish language) Tweets. The hypothesis here is that Bilingual sentiment analysis is more accurate than sentiment analysis in a single language (English or Spanish) when analyzing bilingual tweets. In general, currently sentiment analysis in bilingual tweets is done against an English dictionary. For each of the test cases in this thesis’ experiment we will use the Python NLTK sentiment package.
23
Hu, Jerine, e Jerine Hu. "Sentiment Analysis on Social Media Platforms". Thesis, The University of Arizona, 2017. http://hdl.handle.net/10150/625009.
Testo completoAbstract (sommario):
Sentiment analysis is emerging as a tool that businesses can use to monitor public opinion about their brand. Social media sites such as Twitter provide rich, varying sources of sentiment data to analyze. If social network users conform to sentiments they are exposed to, then businesses can manipulate sentiment on social media to their advantage. In my thesis, I present the codes I developed using Python and Tweepy to gather tweets about the trending topic Standing Rock, explain how sentiment analysis was performed on this data using Semantria, and demonstrate how visualization of sentiment analyses with Tableau can easily illustrate patterns and themes. The results show consistently positive-leaning sentiment among a growing network of users in the absence of an external shock, suggesting that users indeed conform to a convergent sentiment on Twitter. The potential power of sentiment analysis as a business intelligence tool can be applied if firms monitor and analyze social media sentiment to capitalize on existing products and new opportunities.
24
Loreggia, Andrea. "Iterative Voting, Control and Sentiment Analysis". Doctoral thesis, Università degli studi di Padova, 2016. http://hdl.handle.net/11577/3424803.
Testo completoAbstract (sommario):
In multi-agent systems agents often need to take a collective decision based on the preferences of individuals. A voting rule is used to decide which decision to take, mapping the agents' preferences over the possible candidate decisions into a winning decision for the collection of agents. In these kind of scenarios acting strategically can be seen in two opposite way. On one hand it may be desirable that agents do not have any incentive to act strategically. That is, to misreport their preferences in order to influence the result of the voting rule in their favor or acting on the structure of the election to change the outcome. On the other hand manipulation can be used to improve the quality of the outcome by enlarging the consensus of the winner. These two different scenarios are studied in this thesis. The first one by modeling and describing a natural form of control named ``replacement control'' and characterizing for several voting rules its computational complexity. The second scenario is studied in the form of iterative voting frameworks where individuals are allowed to change their preferences to change the outcome of the election. Computational social choice techniques can be used in very different scenarios. This work reports a first attempt to introduce the use of voting procedures in the field of sentiment analysis. In this area computer scientists extract the opinion of the community about a specific item. This opinion is extracted aggregating the opinion expressed by each individual which leaves a text in a blog or social network about the given item. We studied and proposed a new aggregation method which can improve performances of sentiment analysis, this new technique is a new variance of a well-known voting rule called Borda.Nei sistemi multi agente spesso nasce la necessità di prendere decisioni collettive basate sulle preferenze dei singoli individui. A tal fine può essere utilizzata una regola di voto che, aggregando le preferenze dei singoli agenti, trovi una soluzione che rappresenti la collettività. In questi scenari la possibilità di agire in modo strategico può essere vista da due diversi e opposti punti di vista. Da una parte può essere desiderabile che gli agenti non abbiano alcun incentivo ad agire strategicamente, ovvero che gli agenti non abbiano incentivi a riportare in modo scorretto le proprie preferenze per influenzare il risultato dell'elezione a proprio favore, oppure che non agiscano sulla struttura del sistema elettorale stesso per cambiarne il risultato finale. D'altra parte l'azione strategica può essere utilizzata per migliorare la qualità del risultato o per incrementare il consenso del vincitore. Questi due diversi scenari sono studiati ed analizzati nella tesi. Il primo modellando e descrivendo una forma naturale di controllo chiamato "replacement control" descrivendo la complessità computazione di tale azione strategica per diverse regole di voto. Il secondo scenario è studiato nella forma dei sistemi di voto iterativi nei quali i singoli individui hanno la possibilità di cambiare le proprie preferenze al fine di influenzare il risultato dell'elezione. Le tecniche di Computational Social Choice inoltre possono essere usate in diverse situazioni. Il lavoro di tesi riporta un primo tentativo di introdurre l'uso di sistemi elettorali nel campo dell'analisi del sentimento. In questo contesto i ricercatori estraggono le opinioni della comunità riguardanti un particolare elemento di interesse. L'opinione collettiva è estratta aggregando le opinioni espresse dai singoli individui che discutono o parlano dell'elemento di interesse attraverso testi pubblicati in blog o social network. Il lavoro di tesi studia una nuova procedura di aggregazione proponendo una nuova variante di una regola di voto ben conosciuta qual è Borda. Tale nuova procedura di aggregazione migliora le performance dell'analisi del sentimento classica.
25
Mattila, Max, e Hassan Salman. "Analysing Social Media Marketing on Twitter using Sentiment Analysis". Thesis, KTH, Skolan för elektroteknik och datavetenskap (EECS), 2018. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-229787.
Testo completoAbstract (sommario):
Social media is an increasingly important marketing platform in today’s society, and many businesses use them in one way or another in their advertising. This report aimed to determine the effect of different factors on the sentiment in the response to a tweet posted on Twitter for advertising purposes by companies in the fast food sector in North America. The factors considered were the time of posting, the length and the sentiment of a tweet, along with the presence of media other than text in the tweet. Sentiment was extracted from samples of the response to the advertising tweets collected daily between the 27th of March and the 28th of April and plotted against the factors mentioned. The results indicate that the sentiment of the advertising tweet along with the time of posting had the biggest impact on the response, though no definitive conclusions on their effects could be drawn.Sociala medier är en allt viktigare marknadsföringsplattform i dagens samhälle, och många företag använder dem på ett eller annat sätt i sin marknadsföring. Syftet med denna studie är att genom attitydanalys undersöka hur ett antal faktorer inom marknadsföring på det sociala mediet Twitter påverkar responsen till den. Dessa faktorer var följande: inläggets tid, längd och attityd, samt förekomst av media i inlägget. Inläggen samlades från Twitter mellan 28. mars och 28. april och attityden i dem mättes genom attitydanalys, varpå attityden i svaren till reklaminläggen jämfördes baserat på de ovannämnda faktorerna. Resultaten visar på att attityden i reklaminläggen och tiden då de läggs upp har störst påverkan på hur svaren ser ut, men inga säkra slutsatser har kunnat dras.
26
Altrabsheh, Nabeela. "Sentiment analysis on students' real-time feedback". Thesis, University of Portsmouth, 2016. https://researchportal.port.ac.uk/portal/en/theses/sentiment-analysis-on-students-realtime-feedback(4675efcd-c784-44e1-9e2f-47f42d520f7b).html.
Testo completoAbstract (sommario):
Previous literature identifies that students’ real-time feedback is important in the learning process. There are numerous studies that have collected students’ feedback in real time. However, they include several limitations of which the most important is analysing the feedback. In this thesis, we address these limitations by proposing a system that will automatically analyse students’ feedback in real time and present the analysis results to the lecturer. To create such a system, we propose the use of sentiment analysis. The extensive literature highlights the importance of this research in the sentiment analysis field, as there is no research exploring sentiment analysis for students’ real-time feedback and research in the educational domain has been focused mainly on e-learning. The literature shows that emotions are also important in learning and could be detected using sentiment analysis. Consequently, this research is also important as there is little research in detecting emotion from students’ feedback. Polarity detection is explored and two experiments were done to identify the best combination of preprocessing, features and machine learning techniques to create an optimal model for polarity detection of students’ feedback. As a result, we found that the optimal model was to use a lower level of preprocessing, unigrams for features and Complement Naive Bayes for the classifier. Emotion detection in students’ feedback is also explored. Two experiments were done to find the optimal model to detect emotions related to learning from students’ feedback. The results showed that models which detect a single emotion have a better performance than multiple emotion models. Three emotions (i.e. Amused, Bored, and Excited) were more easily detected than others. The optimal model to detect emotion included a low level preprocessing, unigrams as features and Complement Naive Bayes as the classifier. Sarcasm detection in students’ feedback is explored. Our results showed that the lower level of preprocessing led to the best performance. Moreover, by adding other features, such as polarity and emotions, to the unigrams, sarcasm detection increased. Lastly, the best classifier to detect sarcasm was Complement Naive Bayes. Visualisation of the sentiment analysis models results is also investigated. Six visualisations were created and evaluated by lecturers. The findings indicated that there lecturers had no strong preference for a specific visualisation. The system evaluation was performed in real-life settings. As a result, we found that the lecturers were highly satisfied with the system, while the students had a neutral towards negative view of it.
27
Niu, Teng. "Sentiment Analysis on Multi-view Social Data". Thesis, Université d'Ottawa / University of Ottawa, 2016. http://hdl.handle.net/10393/34218.
Testo completoAbstract (sommario):
With the proliferation of social networks, people are likely to share their opinions about news, social events and products on the Web. There is an increasing interest in understanding users’ attitude or sentiment from the large repository of opinion-rich data on the Web. This can benefit many commercial and political applications. Primarily, the researchers concentrated on the documents such as users’ comments on the purchased products. Recent works show that visual appearance also conveys rich human affection that can be predicted. While great efforts have been devoted on the single media, either text or image, little attempts are paid for the joint analysis of multi-view data which is becoming a prevalent form in the social media. For example, paired with the posted textual messages on Twitter, users are likely to upload images and videos which may carry their affective states. One common obstacle is the lack of sufficient manually annotated instances for model learning and performance evaluation. To prompt the researches on this problem, we introduce a multi-view sentiment analysis dataset (MVSA) including a set of manually annotated image-text pairs collected from Twitter. The dataset can be utilized as a valuable benchmark for both single-view and multi-view sentiment analysis. In this thesis, we further conduct a comprehensive study on computational analysis of sentiment from the multi-view data. The state-of-the-art approaches on single view (image or text) or multi view (image and text) data are introduced, and compared through extensive experiments
conducted on our constructed dataset and other public datasets. More importantly, the effectiveness of the correlation between different views is also studied using the widely used fusion strategies and advanced multi-view feature extraction methods.
28
Remus, Robert. "Genre and Domain Dependencies in Sentiment Analysis". Doctoral thesis, Universitätsbibliothek Leipzig, 2015. http://nbn-resolving.de/urn:nbn:de:bsz:15-qucosa-165438.
Testo completoAbstract (sommario):
Genre and domain influence an author\'s style of writing and therefore a text\'s characteristics. Natural language processing is prone to such variations in textual characteristics: it is said to be genre and domain dependent.
This thesis investigates genre and domain dependencies in sentiment analysis. Its goal is to support the development of robust sentiment analysis approaches that work well and in a predictable manner under different conditions, i.e. for different genres and domains.
Initially, we show that a prototypical approach to sentiment analysis -- viz. a supervised machine learning model based on word n-gram features -- performs differently on gold standards that originate from differing genres and domains, but performs similarly on gold standards that originate from resembling genres and domains. We show that these gold standards differ in certain textual characteristics, viz. their domain complexity. We find a strong linear relation between our approach\'s accuracy on a particular gold standard and its domain complexity, which we then use to estimate our approach\'s accuracy.
Subsequently, we use certain textual characteristics -- viz. domain complexity, domain similarity, and readability -- in a variety of applications. Domain complexity and domain similarity measures are used to determine parameter settings in two tasks. Domain complexity guides us in model selection for in-domain polarity classification, viz. in decisions regarding word n-gram model order and word n-gram feature selection. Domain complexity and domain similarity guide us in domain adaptation. We propose a novel domain adaptation scheme and apply it to cross-domain polarity classification in semi- and unsupervised domain adaptation scenarios. Readability is used for feature engineering. We propose to adopt readability gradings, readability indicators as well as word and syntax distributions as features for subjectivity classification.
Moreover, we generalize a framework for modeling and representing negation in machine learning-based sentiment analysis. This framework is applied to in-domain and cross-domain polarity classification. We investigate the relation between implicit and explicit negation modeling, the influence of negation scope detection methods, and the efficiency of the framework in different domains. Finally, we carry out a case study in which we transfer the core methods of our thesis -- viz. domain complexity-based accuracy estimation, domain complexity-based model selection, and negation modeling -- to a gold standard that originates from a genre and domain hitherto not used in this thesis.
29
Mukras, Rahman. "Representation and learning schemes for sentiment analysis". Thesis, Robert Gordon University, 2009. http://hdl.handle.net/10059/379.
Testo completoAbstract (sommario):
This thesis identifies four novel techniques of improving the performance of sentiment analysis of text systems. Thes include feature extraction and selection, enrichment of the document representation and exploitation of the ordinal structure of rating classes. The techniques were evaluated on four sentiment-rich corpora, using two well-known classifiers: Support Vector Machines and Na¨ıve Bayes. This thesis proposes the Part-of-Speech Pattern Selector (PPS), which is a novel technique for automatically selecting Part-of-Speech (PoS) patterns. The PPS selects its patterns from a background dataset by use of a number of measures including Document Frequency, Information Gain, and the Chi-Squared Score. Extensive empirical results show that these patterns perform just as well as the manually selected ones. This has important implications in terms of both the cost and the time spent in manual pattern construction. The position of a phrase within a document is shown to have an influence on its sentiment orientation, and that document classification performance can be improved by weighting phrases in this regard. It is, however, also shown to be necessary to sample the distribution of sentiment rich phrases within documents of a given domain prior to adopting a phrase weighting criteria. A key factor in choosing a classifier for an Ordinal Sentiment Classification (OSC) problem is its ability to address ordinal inter-class similarities. Two types of classifiers are investigated: Those that can inherently solve multi-class problems, and those that decompose a multi-class problem into a sequence of binary problems. Empirical results showed the former to be more effective with regard to both mean squared error and classification time performances. Important features in an OSC problem are shown to distribute themselves across similar classes. Most feature selection techniques are ignorant of inter-class similarities and hence easily overlook such features. The Ordinal Smoothing Procedure (OSP), which augments inter-class similarities into the feature selection process, is introduced in this thesis. Empirical results show the OSP to have a positive effect on mean squared error performance.
30
Zhu, M. (Mo). "Sentiment analysis on medical treatment of depression". Master's thesis, University of Oulu, 2016. http://urn.fi/URN:NBN:fi:oulu-201611103002.
Testo completoAbstract (sommario):
Applying ICT approach to contribute to improving human’s life is a good purpose for researchers (Pannel, 1993). Thus, implementing products or services with new techniques can be quite an interesting and meaningful topic. Nature language processing is a mature technique may apply machine learning and this technique has already been applied to many applications to server people such as Siri, Chatbot and Google Now. Sentiment analysis is a subject in nature language processing, however, it has not been applied to many fields in our daily life.
Depression is a mental disease which caused a lot of trouble in sociality (Hamilton,1960). It causes a huge damage in people’s daily life in the mental aspect instead of the physical painful. And it is also proved by (Help, 2013) that this kind of mental causing a lot of troubles to patients. Quite many researchers are working to find some better treatments for it. However, similar to any other mental disease, there are many treatments existing for different patients and finding the best treatment for a patient can be a quite difficult job. Thus, in this research, I’m trying to validate the function of the sentiment analysis system by applying the data about depression.
In order to achieve this problem, I have defined three research questions which lead to solving this problem. (1) What is the best algorithm for implementing the sentiment analysis system? (2) What is the best existing sentiment lexicon library which can be applied to implement the sentiment analyzing system? (3) How to implement the sentiment analysis system with a selected sentiment lexicon library? During the research process, I review the literature which is related to these questions to find out the answers.
After all, I selected an open source the sentiment lexical library named SentiWordNot3.0 which was implemented based on an algorithm has the feature of Kth-Nearest Neighbor algorithm and Support Vector Machine. And it proved that following approach can be actually used in sentiment analysis in the medical domain.
31
Chalorthorn, Tawunrat. "Quantitative assessment of factors in sentiment analysis". Thesis, Northumbria University, 2016. http://nrl.northumbria.ac.uk/30233/.
Testo completoAbstract (sommario):
Sentiment can be defined as a tendency to experience certain emotions in relation to a particular object or person. Sentiment may be expressed in writing, in which case determining that sentiment algorithmically is known as sentiment analysis. Sentiment analysis is often applied to Internet texts such as product reviews, websites, blogs, or tweets, where automatically determining published feeling towards a product, or service is very useful to marketers or opinion analysts. The main goal of sentiment analysis is to identify the polarity of natural language text. This thesis sets out to examine quantitatively the factors that have an effect on sentiment analysis. The factors that are commonly used in sentiment analysis are text features, sentiment lexica or resources, and the machine learning algorithms employed. The main aim of this thesis is to investigate systematically the interaction between sentiment analysis factors and machine learning algorithms in order to improve sentiment analysis performance as compared to the opinions of human assessors. A software system known as TJP was designed and developed to support this investigation. The research reported here has three main parts. Firstly, the role of data pre-processing was investigated with TJP using a combination of features together with publically available datasets. This considers the relationship and relative importance of superficial text features such as emoticons, n-grams, negations, hashtags, repeated letters, special characters, slang, and stopwords. The resulting statistical analysis suggests that a combination of all of these features achieves better accuracy with the dataset, and had a considerable effect on system performance. Secondly, the effect of human marked up training data was considered, since this is required by supervised machine learning algorithms. The results gained from TJP suggest that training data greatly augments sentiment analysis performance. However, the combination of training data and sentiment lexica seems to provide optimal performance. Nevertheless, one particular sentiment lexicon, AFINN, contributed better than others in the absence of training data, and therefore would be appropriate for unsupervised approaches to sentiment analysis. Finally, the performance of two sophisticated ensemble machine learning algorithms was investigated. Both the Arbiter Tree and Combiner Tree were chosen since neither of them has previously been used with sentiment analysis. The objective here was to demonstrate their applicability and effectiveness compared to that of the leading single machine learning algorithms, Naïve Bayes, and Support Vector Machines. The results showed that whilst either can be applied to sentiment analysis, the Arbiter Tree ensemble algorithm achieved better accuracy performance than either the Combiner Tree or any single machine learning algorithm.
32
Zhang, Jun. "Sentiment analysis of movie reviews in Chinese". Thesis, Uppsala universitet, Institutionen för lingvistik och filologi, 2020. http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-412670.
Testo completoAbstract (sommario):
Sentiment analysis aims at figuring out the opinions of the users towards a certain service or product. In this research, the aim is at classifying the sentiments of users based on the comments they have posed on Douban movie website. In this thesis, I try two different ways to classify the sentiments: with the first one classifying comments into five classes of ratings from 1 to 5, and with the second one classifying comments into three classes of ratings: negative, neutral and positive. For the latter, the ratings of 1 and 2 are grouped as negative, the ratings of 3 neutral and the ratings of 4 and 5 positive. First, Term Frequency Inverse Document Frequency (TF-IDF) is used as the feature extraction technique for machine learning algorithms. Chi Square and Mutual Information are used for feature selection. The selected features are fed into different machine learning methods: Logistic Regression, Linear SVC, SGD classifier and Multinomial Naive Bayes. The performance of models with feature selection will be compared with the performance of models without feature selection for 5-class classification as well as 3-class classification. Also, fastText and Skip-Gram are used as embedding methods for deep learning algorithms LSTM and BILSTM. FastText will also be used for both embedding as well as being a classifier. The aim is to compare different machine learning and deep learning algorithms using different vectorization methods to see which model performs the best regarding both 5-class and 3-class classification. The two classification strategies will be compared with each other in terms of error analysis. The aim is to figure out the similarities and differences of misclassifications made by two different classification strategies.
33
Di, Bari Marilena. "Improving multilingual sentiment analysis using linguistic knowledge". Thesis, University of Leeds, 2015. http://etheses.whiterose.ac.uk/11883/.
Testo completoAbstract (sommario):
The need for the automatic analysis of opinions in written texts, which has been growing in recent years in several domains, has made Sentiment Analysis a very popular field (Liu 2012). In this area, systems have been traditionally classifying sentences as positive or negative only in accordance to the sentiment that words most frequently assume (e.g. “angry” negative, “beautiful” positive). Such strategies present two main limitations: 1. Multiple opinions often appear in the same sentence, with each expressing an opposing sentiment on different subjects (e.g. a positive opinion is expressed on the plot of a film, but a negative one on the actors' performance). 2. The most frequent sentiment, collected in sentiment dictionaries, does not take into account the fact that context often alters the orientation. Sentiment dictionaries have also been demonstrated to have small coverage (Di Bari, Sharoff et al. 2013, Di Bari 2015). As a consequence, I propose an automatic system based on deep linguistic knowledge given in particular by dependency parsing relations (Nivre 2005) and by attributes taken from the Appraisal framework (Martin and White 2005), a theory concerned with the language of evaluation, attitude and emotion within Systemic Functional Linguistics (Halliday 1978). As a basis for the creation of the automatic system, I tailored an annotation scheme called SentiML inspired by previous works (Whitelaw, Garg et al. 2005, Bloom, Garg et al. 2007, Bloom and Argamon 2009) and carried out the annotation task in three languages (English, Italian and Russian) by using MAE (Stubbs 2011). The resulting corpora consist of around 500 sentences and 9000 tokens for each language. The corpora contain both original texts and translations of different types: news, political speeches and TED talks (Cettolo, Girardi et al. 2012). The foundation of SentiML lies in the fact that an opinion can be captured in a pair consisting of usually two words with different functions: a target as the expression the sentiment refers to, and a modifier as the expression conveying the sentiment. The pair consisting of the target and the modifier altogether is called appraisal group. Along with these main categories, the annotation includes their attributes, among which the most important are the appraisal type according to the Appraisal framework (‘affect’, ‘appreciation’, ‘judgement’) and the orientation (‘positive’ or ‘negative’, both out-of-context and contextual). A detailed manual analysis of the translation strategies (Baker 2002) and the appraisal types across the corpora, supported by insights from Corpus Linguistics has been carried out. The most interesting expressions found during such analysis have been automatically analysed afterwards with the aim of having a further evaluation of the system. Nonetheless, the main evaluation consists of a comparison with a rule-based system that makes use of already existing tools such as the part-of-speech (POS) tagger and the sentiment dictionary. The main objective of this work is to demonstrate that the Appraisal framework and Sentiment analysis can successfully support each other. The additional consideration that this has been done not only for English, but in parallel for Italian and Russian (and as one of the first applications of the Appraisal Framework in these languages) and for different text types, makes the research unique. Moreover, because the methodology used to compare a variety of linguistic features (morphological, grammatical, lexical, syntactical) at work in sentiment analysis has been applied to three languages belonging to different families (Germanic, Romance and Slavonic), it is expected to be generalizable to other languages. As far as the practical applications are concerned, the automatic system could be used in any field in which written opinions need to be analysed. In the meanwhile, the new individual resources such as the annotated corpora and the Maltparser models for Italian and Russian have been made publicly available.
34
Zimbra, David. "Stakeholder and Sentiment Analysis in Web Forums". Diss., The University of Arizona, 2012. http://hdl.handle.net/10150/222894.
Testo completoAbstract (sommario):
Web forums offer open and interactive social communication platforms for numerous participants to share information and offer perspectives on a variety of business and social issues with audiences around the world. In addition to facilitating widespread communication, these web forums contain massive amounts of data and represent rich sources of information that can be utilized to advance the understanding of participants and society. In particular, web forums pertaining to firms and their customers, employees, and investors, represent valuable resources for the acquisition of business intelligence. However, web forums represent a complex analytic landscape requiring the development of automated, intelligent, and scalable analytic approaches. The dissertation follows the design science paradigm in management information systems research, and aims to develop and refine approaches to the analysis of web forums, and to apply these analytic approaches to firm-related web forums to derive information that may explain and predict firm stock behavior. The designs of the devised approaches to web forum analysis are informed by the stakeholder theory of the firm, and systemic functional linguistic theory. We introduce and advance a stakeholder approach to the analysis of firm-related web forums, and improve existing approaches to sentiment analysis in web forums. In Chapter 2 we develop and deploy a stakeholder framework to analyze a popular firm-related finance web forum and apply the extracted measures to explain firm stock return, volatility, and trading volume. In Chapter 3 we advance the stakeholder framework and perform dynamic analyses of web forums over time, and compare several feature representations of stakeholders and approaches to sentiment analysis. We deploy the stakeholder framework to analyze several firm-related web forums, and apply the derived measures to predict firm stock return and perform simulated trading of firm stock over a one year period to determine the economic value of the extracted information. Finally, in Chapter 4 we develop approaches to improve the scalability of sentiment analysis across multiple web forums in a collection. Overall the dissertation contributes to the literature on the analysis of web forums, and demonstrates the value of firm-related web forums as sources of business intelligence.
35
Xue, Wei. "Aspect Based Sentiment Analysis On Review Data". FIU Digital Commons, 2017. https://digitalcommons.fiu.edu/etd/3721.
Testo completoAbstract (sommario):
With proliferation of user-generated reviews, new opportunities and challenges arise. The advance of Web technologies allows people to access a large amount of reviews of products and services online. Knowing what others like and dislike becomes increasingly important for their decision making in online shopping. The retailers also care more than ever about online reviews, because a vast pool of reviews enables them to monitor reputations and collect feedbacks efficiently. However, people often find difficult times in identifying and summarizing fine-grained sentiments buried in the opinion-rich resources. The traditional sentiment analysis, which focuses on the overall sentiments, fails to uncover the sentiments with regard to the aspects of the reviewed entities.
This dissertation studied the research problem of Aspect Based Sentiment Analysis (ABSA), which is to reveal the aspect-dependent sentiment information of review text. ABSA consists of several subtasks: 1) aspect extraction, 2) aspect term extraction, 3) aspect category classification, and 4) sentiment polarity classification at aspect level. We focused on the approach of topic models and neural networks for ABSA. First, to extract the aspects from a collection of reviews and to detect the sentiment polarity regarding the aspects in each review, we proposed a few probabilistic graphical models, which can model words distribution in reviews and aspect ratings at the same time. Second, we presented a multi-task learning model based on long-short term memory and convolutional neural network for aspect category classification and aspect term extraction. Third, for aspect-level sentiment polarity classification, we developed a gated convolution neural network, which can be applied to aspect category sentiment analysis as well as aspect target sentiment analysis.
36
CAPUA, M. DI. "A DEEP LEARNING APPROACH FOR SENTIMENT ANALYSIS". Doctoral thesis, Università degli Studi di Milano, 2017. http://hdl.handle.net/2434/467844.
Testo completoAbstract (sommario):
La Sentiment Analysis si riferisce alla analisi qualitativa volta ad identificare e classificare opinioni contenute in frasi e testi, allo scopo di stabilire lo “stato d’animo” dell’autore rispetto ad un particolare argomento o prodotto, e di determinare se tale stato è di fatto positivo, negativo oppure neutrale. Le opinioni espresse in un testo, come ad esempio giudizi, sentimenti ed emozioni, sono di recente diventate oggetto di studio e di ricerca sia in ambito accademico che industriale. Sfortunatamente la comprensione del linguaggio, applicata a commenti di utenti, è un attività estremamente complessa per una macchina, specialmente se ci si riferisce ai contesti dei moderni social network. Le modalità in cui le persone si esprimono in linguaggio naturale, sono molteplici, e l’utilizzo “informale” della lingua adottato tipicamente nei social netowrks, genera frasi spesso dense di errori, modi di dire (slang), costrutti sintattici ”personalizzati”, o anche frasi arricchite da caratteri speciali (come l’hashtag in Twitter), il che complica notevolmente l’analisi. Recentemente, le tecniche di Deep Learning, stanno emergendo nel panorama del machine learning, come un modello computazionale che può essere adoperato con efficacia per scoprire relazioni semantiche complesse, all’interno di un testo, anche senza la necessità di dover individuare a priori caratteristiche (features) di tali relazioni. Questi approcci hanno migliorato l’attuale stato dell’arte in diversi settori della Sentiment Analysis, come ad esempio la classificazione di frasi o di documenti, l’apprendimento basato su lexicon, fino ad arrivare alla analisi di fenomeni complessi come il cyber bullismo.
I contributi di questa tesi sono di due tipi. Il primo contributo fornito, relativo ad aspetti generali di Sentiment Analysis, riguarda la proposta di un modello di rete neurale semi supervisionata, basato sulle reti di tipo Deep Belief, in grado di affrontare l’incertezza dei dati insita nelle frasi testuali, con particolare riferimento alla lingua italiana. Il modello proposto è stato testato rispetto a diversi datasets presi dalla letteratura di riferimento, composti da testi relativi a critiche cinematografiche, adottando una rappresentazione dell’informazione basata su vettori (Word2Vec) ed introducendo anche metodi derivati dal campo del Natural Language Processing (NLP). Il secondo contributo fornito in questa tesi, partendo dall’assunto che il cyber bullismo può essere considerato come un caso particolare di Sentiment Analysis, propone un approccio non supervisionato alla rilevazione automatica di tracce di cyber bullismo all’interno di social networks, basato sia su di una rete neurale di tipo GHSOM (Growing Hierarchical Self Organizing Map), sia su di un modello di caratteristiche (features) predefinito. Il modello non supervisionato proposto dimostra di raggiungere comunque risultati interessanti rispetto ai tipici modelli supervisionati, applicati solitamente in questo ambito.Sentiment Analysis refers to the process of computationally identifying and categorizing opinions expressed in a piece of text, in order to determine whether the writer’s attitude towards a particular topic or product is positive, negative, or even neutral. The views expressed and its related concepts, such as feelings, judgments, and emotions have become recently a subject of study and research in both academic and industrial areas. Unfortunately language comprehension of user comments, especially in social networks, is inherently complex to computers. The ways in which humans express themselves with natural language are nearly unlimited and informal texts is riddled with typos, misspellings, badly set up syntactic constructions and also specific symbols (e.g. hashtags in Twitter) which exponentially complicate this task. Recently, deep learning approaches are emerging as powerful computational models that discover intricate semantic representations of texts automatically from data without hand-made feature engineering. These approaches have improved the state-of-the-art in many Sentiment Analysis tasks including sentiment classification of sentences or documents, sentiment lexicon learning and also in more complex problems as cyber bullying detection. The contributions of this work are twofold. First, related to the general Sentiment Analysis problem, we propose a semi-supervised neural network model, based on Deep Belief Networks, able to deal with data uncertainty for text sentences in Italian language. We test this model against some datasets from literature related to movie reviews, adopting a vectorized representation of text (Word2Vec) and exploiting methods from Natural Language Processing (NLP) pre-processing. Second, assuming that the cyber bullying phenomenon can be treated as a particular Sentiment Analysis problem, we propose an unsupervised approach to automatic cyber bullying detection in social networks, based both on Growing Hierarchical Self Organizing Map (GHSOM) and on a new specific features model, showing that our solution can achieve interesting results, respect to classical supervised approaches.
37
Johansson, Henrik, e Anton Lilja. "Method performance difference of sentiment analysis on social media databases : Sentiment classification in social media". Thesis, KTH, Skolan för datavetenskap och kommunikation (CSC), 2016. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-187259.
Testo completoAbstract (sommario):
As the amount of available data have exploded with the in- crease in use of social media the interest of doing sentiment anlysis have increased. However as the source and nature of the data have changed it is possible that the known meth- ods will not perform as before. The purpose of this paper is to examine if such a di erence exist and if the methods can be improved through preprocessing the data. The results show that there is a di erence and that on this new type of data a lexicon approach may be a better choice than a machine learning based one. Preprocessing the data give some but no large improvements.Den explosion av tillgänglig data i och med den ökade an- vändningen av sociala medier har ökat intresset för att göra sentimentsanalys. Men eftersom källan och innehållet för den data som analyseras har förändrats är det möjligt att de metoder som används kommer att prestera annorlunda. Syftet med denna studie är att undersöka om en sådan skill- nad finns och om metodernas trä säkerhet kan ökas genom att förarbeta data. Resultatet visar att det finns en skillnad och att en lexikal analys kan vara ett bättre tillvägagångs- sätt än en metod baserad på maskininlärning. Att förarbeta data visar viss men inte i sammanhanget stor förbättring av resultatet.
38
LYSEDAL, TOMAS. "Sentiment analysis of Swedish social media : Using random indexing to improve cross-domain sentiment classification". Thesis, KTH, Skolan för datavetenskap och kommunikation (CSC), 2014. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-156249.
Testo completoAbstract (sommario):
Social media has grown extremely fast in recent years andin the vast number of posts being made everyday people expresstheir opinions about all kinds of topics. These opinionsare very valuable and there is a need for a way toautomatically identify and extract them. This is what sentimentanalysis is about but there are a number of issuesrelated to this task. In particular the large number anddiversity of the texts to analyze causes problems for ordinarymethods of natural language processing. In this thesisa method utilizing a technique called Random Indexing isproposed which tries to overcome some of the issues. Theconclusion is that the use of Random Indexing does aid insolving the problem but also that more work is needed inorder to have a fully satisfying solution.
39
Haider, Syed Zeeshan. "AN ONTOLOGY BASED SENTIMENT ANALYSIS : A Case Study". Thesis, Högskolan i Skövde, Institutionen för kommunikation och information, 2012. http://urn.kb.se/resolve?urn=urn:nbn:se:his:diva-6387.
Testo completoAbstract (sommario):
Business through e-commerce has become popular recently due to the massive amount of information available on internet. This has resulted in the abnormal number of reviews on websites like www.amazon.com and www.ebay.com, where customers express their opinions about the purchases they have made. Analyzing customer’s behavior has become very important for the organizations to find new market trends and insights. For the potential customer it becomes really difficult to get the knowledge about a product in the presence of such huge number of reviews and to sort the useful reviews and make good decision. The reviews available on these websites are in heterogeneous form i.e. structured and unstructured form and needs to be stored in a consistent format. Since good decision requires quality information in limited amount of time, Yaakub et, al.(2011) have proposed an ontology that uses a multidimensional model to integrate customer’s characteristics and their comments about products. This approach first identifies the entities and then sentiments present in the customers reviews related to mobiles are transformed into an attribute table by using a 7 point polarity system (-3 to 3). The research proposed by Yaakub et, al.(2011) is in developing stage. The limitation of their approach is that the ontology proposed by them is too general. The authors have shown their desire that it should be tested for a large group of products. Also, Yaakub et, al.(2011) have used very short and simple comments for the manual extraction of features for which a sentiment has been expressed. Usually comments present on e-commerce websites are not that short and simple. In order to fulfill the aim of this thesis project, a case study has been conducted on websites www.amazon.com and www.ebay.com and the ontology proposed by Yaakub et, al.(2011) has been refined for the three categories of mobile phones: smart phones, wet and dirty mobile phones and simple mobile phones. Further, sentiment analysis has been conducted by first using the ontology proposed by Yaakub et, al.(2011) and then by using the refined version of the ontologies for the three categories of mobile in order to compare the results.
40
Duarte, Eduardo Santos. "Sentiment analysis on twitter for the portuguese language". Master's thesis, Faculdade de Ciências e Tecnologia, 2013. http://hdl.handle.net/10362/11338.
Testo completoAbstract (sommario):
Dissertação para obtenção do Grau de Mestre em
Engenharia InformáticaWith the growth and popularity of the internet and more specifically of social networks, users can more easily share their thoughts, insights and experiences with others. Messages shared via social networks provide useful information for several applications, such as monitoring specific targets for sentiment or comparing the public sentiment on several targets, avoiding the traditional marketing research method with the use of surveys to explicitly get the public opinion. To extract information from the large amounts of messages that are shared, it is best to use an automated program to process these messages. Sentiment analysis is an automated process to determine the sentiment expressed in natural language in text. Sentiment is a broad term, but here we are focussed in opinions and emotions that are expressed in text. Nowadays, out of the existing social network websites, Twitter is considered the best one for this kind of analysis. Twitter allows users to share their opinion on several topics and entities, by means of short messages. The messages may be malformed and contain spelling errors, therefore some treatment of the text may be necessary before the analysis, such as spell checks. To know what the message is focusing on it is necessary to find these entities on the text such as people, locations, organizations, products, etc. and then analyse the rest of the text and obtain what is said about that specific entity. With the analysis of several messages, we can have a general idea on what the public thinks regarding many different entities. It is our goal to extract as much information concerning different entities from tweets in the Portuguese language. Here it is shown different techniques that may be used as well as examples and results on state-of-the-art related work. Using a semantic approach, from these messages we were able to find and extract named entities and assigning sentiment values for each found entity, producing a complete tool competitive with existing solutions. The sentiment classification and assigning to entities is based on the grammatical construction of the message. These results are then used to be viewed by the user in real time or stored to be viewed latter. This analysis provides ways to view and compare the public sentiment regarding these entities, showing the favourite brands, companies and people, as well as showing the growth of the sentiment over time.
41
Muhammad, Aminu. "Contextual lexicon-based sentiment analysis for social media". Thesis, Robert Gordon University, 2016. http://hdl.handle.net/10059/1571.
Testo completoAbstract (sommario):
Sentiment analysis concerns the computational study of opinions expressed in text. Social media domains provide a wealth of opinionated data, thus, creating a greater need for sentiment analysis. Typically, sentiment lexicons that capture term-sentiment association knowledge are commonly used to develop sentiment analysis systems. However, the nature of social media content calls for analysis methods and knowledge sources that are better able to adapt to changing vocabulary. Invariably existing sentiment lexicon knowledge cannot usefully handle social media vocabulary which is typically informal and changeable yet rich in sentiment. This, in turn, has implications on the analyser's ability to effectively capture the context therein and to interpret the sentiment polarity from the lexicons. In this thesis we use SentiWordNet, a popular sentiment-rich lexicon with a substantial vocabulary coverage and explore how to adapt it for social media sentiment analysis. Firstly, the thesis identifies a set of strategies to incorporate the effect of modifiers on sentiment-bearing terms (local context). These modifiers include: contextual valence shifters, non-lexical sentiment modifiers typical in social media and discourse structures. Secondly, the thesis introduces an approach in which a domain-specific lexicon is generated using a distant supervision method and integrated with a general-purpose lexicon, using a weighted strategy, to form a hybrid (domain-adapted) lexicon. This has the dual purpose of enriching term coverage of the general purpose lexicon with non-standard but sentiment-rich terms as well as adjusting sentiment semantics of terms. Here, we identified two term-sentiment association metrics based on Term Frequency and Inverse Document Frequency that are able to outperform the state-of-the-art Point-wise Mutual Information on social media data. As distant supervision may not be readily applicable on some social media domains, we explore the cross-domain transferability of a hybrid lexicon. Thirdly, we introduce an approach for improving distant-supervised sentiment classification with knowledge from local context analysis, domain-adapted (hybrid) and emotion lexicons. Finally, we conduct a comprehensive evaluation of all identified approaches using six sentiment-rich social media datasets.
42
Barnes, Jeremy. "Cross-lingual sentiment analysis for under-resourced languages". Doctoral thesis, Universitat Pompeu Fabra, 2019. http://hdl.handle.net/10803/665480.
Testo completoAbstract (sommario):
Sentiment Analysis is a task that aims to calculate the polarity of text automatically. While some languages, such as English, have a vast array of resources to enable sentiment analysis, most under-resourced languages lack them. Cross-lingual Sentiment Analysis (CLSA) attempts to make use of resource-rich languages in order to create or improve sentiment analysis systems in an under-resourced language. In this thesis, we propose cross-lingual sentiment approaches that have minimal parallel data requirements, while making the best use of available monolingual data. We propose a model to incorporate sentiment information into bilingual distributional representations, by jointly optimizing them for semantics and sentiment, showing state-of-the-art performance when combined with machine translation. We then move these approaches to aspect-level and subsequently test them on a variety of language families and domains. Finally, we show that this approach can also be suitable for domain adaptation.L’anàlisi de sentiment és una tasca que ens permet calcular la polaritat de un text de manera automàtica. Mentre algunes llengües, com l’anglès per exemple, tenen una àmplia varietat de recursos per crear sistemes d’anàlisi de sentiment, n’hi ha més que els troben a faltar. L’Anàlisi de Sentiment Cross-lingüe (ASCL) intenta fer servir els recursos de llengües riques en recursos per crear o millorar sistemes d’anàlisi de sentiment en llengües pobres en recursos. A aquesta tesi proposem mètodes d’anàlisi de sentiment cross-lingües que requereixen menys data paral·lela i treuen el màxim profit de data monolingüe que tenim a l’abast. Proposem un model que optimitza les representacions distribucionals cross-lingües perquè tinguin informació semàntica i també de sentiment, i que demostra ser l’estat de l’art en combinant-se amb traducció automàtica. Després passem a un nivell de granularitat més fina i examinem com canvia el rendiment dels models amb diferents llengües metes i dominis. Finalment, demostrem que aquestes tècniques també són adequats per a l’adaptació de domini.
43
Dettori, Emilio. "Sentiment Analysis per la moderazione di una community". Bachelor's thesis, Alma Mater Studiorum - Università di Bologna, 2018.
Cerca il testo completoAbstract (sommario):
Questa tesi, dopo aver illustrato i concetti di Machine Learning e Natural Language Processing, descrive il processo di Sentiment Analysis e le sue applicazioni. L'obiettivo del progetto di tesi è stato quello di studiare un sistema di moderazione automatica testuale di una community online, svolto utilizzando le tre tecniche descritte. In particolare, il fine del progetto è quello di effettuare un'analisi lessicale del testo su un corpus creato appositamente, per poi sviluppare algoritmi di Machine Learning in grado di apprendere da esso. Per ogni tecnologia analizzata sono mostrati esempi e casi d'uso.
44
Tramontano, Matteo. "sentiment analysis: previsione con tecniche di intelligenza artificiale". Bachelor's thesis, Alma Mater Studiorum - Università di Bologna, 2020. http://amslaurea.unibo.it/20387/.
Testo completoAbstract (sommario):
La tesi tratta l'argomento chiamato Sentiment Analysis, ovvero un campo dell'elaborazione del linguaggio naturale che si occupa di costruire sistemi per l'identificazione ed estrazione di opinioni dal testo basandosi sui principali metodi di linguistica computazionale e di analisi testuale.
In particolare ha lo scopo di mostrare realizzazione ed applicazione di questa pratica focalizzandosi sugli algoritmi che la costituiscono.
Viene anche analizzato il suo utilizzo al fine di effettuare previsioni nel futuro con particolare attenzione all'ambito economico-finanziario.
45
Torchi, Andrea. "Sperimentazioni per "Sentiment Analysis" tramite Reti Neurali Profonde". Master's thesis, Alma Mater Studiorum - Università di Bologna, 2020.
Cerca il testo completoAbstract (sommario):
Il lavoro di questa tesi verte sulla progettazione e sulla realizzazione di un software di machine-learning che svolga un compito di Sentiment Analysis, in particolar modo è stato svolto nell'ambito di un progetto commissionato all'azienda presso la quale ho svolto il mio tirocinio.
Come fonte di dati da cui iniziare e su cui basare il progetto ho scelto alcuni social networks, per via della grande quantità di dati che offrono.
Prima e durante il lavoro di tesi ho studiato i principi del machine-learning ed in particolare delle reti neurali, concetti che furono in seguito applicati nella realizzazione delle reti per lo svolgimento del compito di Sentiment Analysis.
46
Ziegelmayer, Dominique [Verfasser]. "Character n-gram-based sentiment analysis / Dominique Ziegelmayer". München : Verlag Dr. Hut, 2015. http://d-nb.info/1060587688/34.
Testo completo
47
Stenqvist, Evita, e Jacob Lönnö. "Predicting Bitcoin price fluctuation with Twitter sentiment analysis". Thesis, KTH, Skolan för datavetenskap och kommunikation (CSC), 2017. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-209191.
Testo completoAbstract (sommario):
Programmatically deriving sentiment has been the topic of many a thesis: it’s application in analyzing 140 character sentences, to that of 400-word Hemingway sentences; the methods ranging from naive rule based checks, to deeply layered neural networks. Unsurprisingly, sentiment analysis has been used to gain useful insight across industries, most notably in digital marketing and financial analysis. An advancement seemingly more excitable to the mainstream, Bitcoin, has risen in number of Google searches by three-folds since the beginning of this year alone, not unlike it’s exchange rate. The decentralized cryptocurrency, arguably, by design, a pure free market commodity – and as such, public perception bears the weight in Bitcoins monetary valuation. This thesis looks toward these public perceptions, by analyzing 2.27 million Bitcoin-related tweets for sentiment fluctuations that could indicate a price change in the near future. This is done by a naive method of solely attributing rise or fall based on the severity of aggregated Twitter sentiment change over periods ranging between 5 minutes and 4 hours, and then shifting these predictions forward in time 1, 2, 3 or 4 time periods to indicate the corresponding BTC interval time. The prediction model evaluation showed that aggregating tweet sentiments over a 30 min period with 4 shifts forward, and a sentiment change threshold of 2.2%, yielded a 79% accuracy.Ämnet sentiment analysis, att programmatiskt härleda underliggande känslor i text, ligger som grund för många avhandlingar: hur det tillämpas bäst på 140 teckens meningar såväl som på 400-ords meningar a’la Hemingway, metoderna sträcker sig ifrån naiva, regelbaserade, till neurala nätverk. Givetvis sträcker sig intresset för sentiment analys utanför forskningsvärlden för att ta fram insikter i en rad branscher, men framförallt i digital marknadsföring och financiell analys. Sedan början på året har den digitala valutan Bitcoin stigit trefaldigt i sökningar på Google, likt priset på valutan. Då Bitcoins decentraliserade design är helt transparant och oreglerad, verkar den under ideala marknadsekonomiska förutsättningar. På så vis regleras Bitcoins monetära värde av marknadens uppfattning av värdet. Denna avhandling tittar på hur offentliga uppfattningar påverkar Bitcoins pris. Genom att analysera 2,27 miljoner Bitcoin-relaterade tweets för sentiment ändringar, föutspåddes ändringar i Bitcoins pris under begränsade förhållningar. Priset förespåddes att gå upp eller ner beroende på graden av sentiment ändring under en tidsperiod, de testade tidsperioderna låg emellan 5 minuter till 4 timmar. Om en förutspånning görs för en tidsperiod, prövas den emot 1, 2, 3 och 4 skiftningar framåt i tiden för att ange förutspådd Bitcoin pris interval. Utvärderingen av förutspåningar visade att aggregerade tweet-sentiment över en 30-minutersperiod med 4 skift framåt och ett tröskelvärde för förändring av sentimentet på 2,2 % gav ett resultat med 79 % noggrannhet.
48
Wang, Szu-Hung, e 王斯泓. "Sentiment-Guided Attention Mechanism for Sentiment Analysis". Thesis, 2019. http://ndltd.ncl.edu.tw/handle/vz6j9g.
Testo completoAbstract (sommario):
碩士國立臺灣大學
資訊網路與多媒體研究所
107
Sentiment analysis is an important task, which extracts sentiment, emotion or affect in text. The problem is often treated as a classification problem for which deep neural methods have been well explored and attention mechanisms have generated promising performance. Studies have shown that lexicon is highly effective for sentiment analysis. However, lexicon has not been fully utilized by the previous methods. No existing method integrates lexicon into the attention mechanism effectively to solve the problem. This thesis explores the sentiment-guided attention mechanism, which integrates lexicon into attention mechanism and proposes two approaches. First, to utilize sentiment lexicons, we transform lexicon values into guiding weights to minimize the error of attention weights. Second, we propose sentiment multi-head attention to help the model jointly attend to sentiment information provided by the transformed lexicon values. Experiments show that our models outperform state-of-the-art models on six sentiment analysis benchmarks with improved accuracy of 0.12% to 8.12%.
49
Tavares, Cátia Daniela Lopes. "Sentiment analysis to predict the Portuguese economic sentiment based on economic news". Master's thesis, 2021. http://hdl.handle.net/10071/24130.
Testo completoAbstract (sommario):
Measuring the economic sentiment of a country is crucial to understand and predict its
short-term economic condition. This work proposes an automatic sentiment indicator, derived
from collected economic news texts, that is able to accurately measure the current
economic sentiment in Portugal and is highly correlated with the official Economic Sentiment
Indicator, published a few weeks later by the European Commission, based on surveys.
The data used in these experiments consists of almost 90 thousand Portuguese economic
news, extracted from two well-known Portuguese newspapers, covering the period
from 2010 to 2020. Each document was automatically classified with the corresponding
sentiment polarity, using a rule-based approach that proved suitable for detecting the sentiment
in Portuguese economic news. In order to perform sentiment analysis of economic
news, we have also evaluated the adaptation of existing pre-trained modules and performed
experiments with a set of Machine Learning approaches. Experimental results show that
our rule-based approach, that uses manually written rules specific to the economic context,
achieves the best results for automatically detecting the polarity of economic news, largely
surpassing the other approaches.
Our experimental results shows that the sentiment expressed through economic news
constitute a promising way of predicting the economic sentiment, thus allowing to understand
the economic situation in Portugal in almost real time. The developed indicator, based
on the news, give us a predictive power of the economic fluctuations and the sentiment concerning
the economic agents about the present and the future of the economy.Medir o sentimento económico de um país é crucial para compreender e prever a sua condição económica de curto prazo. Este projeto propõe um indicador de sentimento automático, baseado em textos recolhidos de notícias económicas, que é capaz de medir com precisão o sentimento económico atual em Portugal e está altamente correlacionado com o Indicador de Sentimento Económico oficial, publicado pela Comissão Europeia algumas semanas depois e calculado com base em inquéritos. Os dados utilizados nestas experiências consistem em cerca de 90 mil notícias económicas portuguesas, extraídas de dois jornais portugueses de renome, abrangendo o período de 2010 a 2020. Cada notícia foi automaticamente classificada com a polaridade de sentimento que tem associada, através de uma abordagem baseada em regras que provou ser adequada para detectar o sentimento das notícias económicas portuguesas. Para realizar a análise de sentimento das notícias económicas, também avaliámos a adaptação de módulos prétreinados existentes e realizamos experiências com um conjunto de abordagens de Aprendizagem Automática. Resultados experimentais mostram que a nossa abordagem baseada em regras, que usa regras escritas manualmente específicas para o contexto económico, alcança os melhores resultados para detectar automaticamente a polaridade das notícias económicas, superando amplamente as outras abordagens. O nosso estudo mostra que o sentimento expresso através das notícias económicas constitui uma forma promissora de prever o sentimento económico, permitindo entender a situação económica em Portugal quase em tempo real. O indicador desenvolvido, com base nas notícias, tem poder preditivo das flutuações económicas e do sentimento dos agentes económicos acerca do presente e o futuro da economia.
50
Coelho, Pedro Samuel Amaro. "Multi-Topic Sentiment Analysis". Master's thesis, 2013. https://repositorio-aberto.up.pt/handle/10216/69498.
Testo completo