Dissertations / Theses on the topic 'Language of social media'

Statistics show that the world wide web is dominated by a few widely spoken languages. However, in quieter corners of the web, clusters of minority language speakers can be found interacting and sharing content. This study is the first to compare three such clusters of Irish language social media users. Social network analysis of the most active public sites of interaction through Irish – the Irish language blogosphere, the Irish language Twittersphere and a popular Irish language Facebook group – reveals unique networks of individuals communicating through Irish in unique and innovative ways. Firstly, it describes the members and their activity, and the size and structure of the networks they share. Then through focused discourse analysis of the core prolific users in each network it describes how the language has been adapted to computer-mediated communication. This study found that the largest networks of Irish speakers comprised between 150-300 regular participants each. Most members were adults, male, and lived in towns and cities outside of the language’s traditional heartland. Moreover, each group shared one common trait: though scattered geographically, through regular online interaction between core members they behave like communities. They were found to have shared histories, norms and customs, and self-awareness that their groups were unique. Furthermore, core users had adapted the language in new and innovative ways through their online discourse. This study is the first comprehensive audit of who is using the Irish language socially on the web, where they are forming networks online, and how they are adapting the language to online discourse. It makes a unique contribution in re-imagining what constitutes an Irish language community in the context of the Network Society. In the process, it contributes to the growing body of sociolinguistic research into globalisation and local identity on the web.

[ES] En los últimos años, los modelos de aprendizaje automático profundo (AP) han revolucionado los sistemas de procesamiento de lenguaje natural (PLN). Hemos sido testigos de un avance formidable en las capacidades de estos sistemas y actualmente podemos encontrar sistemas que integran modelos PLN de manera ubicua. Algunos ejemplos de estos modelos con los que interaccionamos a diario incluyen modelos que determinan la intención de la persona que escribió un texto, el sentimiento que pretende comunicar un tweet o nuestra ideología política a partir de lo que compartimos en redes sociales. En esta tesis se han propuestos distintos modelos de PNL que abordan tareas que estudian el texto que se comparte en redes sociales. En concreto, este trabajo se centra en dos tareas fundamentalmente: el análisis de sentimientos y el reconocimiento de la personalidad de la persona autora de un texto. La tarea de analizar el sentimiento expresado en un texto es uno de los problemas principales en el PNL y consiste en determinar la polaridad que un texto pretende comunicar. Se trata por lo tanto de una tarea estudiada en profundidad de la cual disponemos de una vasta cantidad de recursos y modelos. Por el contrario, el problema del reconocimiento de personalidad es una tarea revolucionaria que tiene como objetivo determinar la personalidad de los usuarios considerando su estilo de escritura. El estudio de esta tarea es más marginal por lo que disponemos de menos recursos para abordarla pero que no obstante presenta un gran potencial. A pesar de que el enfoque principal de este trabajo fue el desarrollo de modelos de aprendizaje profundo, también hemos propuesto modelos basados en recursos lingüísticos y modelos clásicos del aprendizaje automático. Estos últimos modelos nos han permitido explorar las sutilezas de distintos elementos lingüísticos como por ejemplo el impacto que tienen las emociones en la clasificación correcta del sentimiento expresado en un texto. Posteriormente, tras estos trabajos iniciales se desarrollaron modelos AP, en particular, Redes neuronales convolucionales (RNC) que fueron aplicadas a las tareas previamente citadas. En el caso del reconocimiento de la personalidad, se han comparado modelos clásicos del aprendizaje automático con modelos de aprendizaje profundo, pudiendo establecer una comparativa bajo las mismas premisas. Cabe destacar que el PNL ha evolucionado drásticamente en los últimos años gracias al desarrollo de campañas de evaluación pública, donde múltiples equipos de investigación comparan las capacidades de los modelos que proponen en las mismas condiciones. La mayoría de los modelos presentados en esta tesis fueron o bien evaluados mediante campañas de evaluación públicas, o bien emplearon la configuración de una campaña pública previamente celebrada. Siendo conscientes, por lo tanto, de la importancia de estas campañas para el avance del PNL, desarrollamos una campaña de evaluación pública cuyo objetivo era clasificar el tema tratado en un tweet, para lo cual recogimos y etiquetamos un nuevo conjunto de datos. A medida que avanzabamos en el desarrollo del trabajo de esta tesis, decidimos estudiar en profundidad como las RNC se aplicaban a las tareas de PNL. En este sentido, se exploraron dos líneas de trabajo. En primer lugar, propusimos un método de relleno semántico para RNC, que plantea una nueva manera de representar el texto para resolver tareas de PNL. Y en segundo lugar, se introdujo un marco teórico para abordar una de las críticas más frecuentes del aprendizaje profundo, el cual es la falta de interpretabilidad. Este marco busca visualizar qué patrones léxicos, si los hay, han sido aprendidos por la red para clasificar un texto.
[CA] En els últims anys, els models d'aprenentatge automàtic profund (AP) han revolucionat els sistemes de processament de llenguatge natural (PLN). Hem estat testimonis d'un avanç formidable en les capacitats d'aquests sistemes i actualment podem trobar sistemes que integren models PLN de manera ubiqua. Alguns exemples d'aquests models amb els quals interaccionem diàriament inclouen models que determinen la intenció de la persona que va escriure un text, el sentiment que pretén comunicar un tweet o la nostra ideologia política a partir del que compartim en xarxes socials. En aquesta tesi s'han proposats diferents models de PNL que aborden tasques que estudien el text que es comparteix en xarxes socials. En concret, aquest treball se centra en dues tasques fonamentalment: l'anàlisi de sentiments i el reconeixement de la personalitat de la persona autora d'un text. La tasca d'analitzar el sentiment expressat en un text és un dels problemes principals en el PNL i consisteix a determinar la polaritat que un text pretén comunicar. Es tracta per tant d'una tasca estudiada en profunditat de la qual disposem d'una vasta quantitat de recursos i models. Per contra, el problema del reconeixement de la personalitat és una tasca revolucionària que té com a objectiu determinar la personalitat dels usuaris considerant el seu estil d'escriptura. L'estudi d'aquesta tasca és més marginal i en conseqüència disposem de menys recursos per abordar-la però no obstant i això presenta un gran potencial. Tot i que el fouc principal d'aquest treball va ser el desenvolupament de models d'aprenentatge profund, també hem proposat models basats en recursos lingüístics i models clàssics de l'aprenentatge automàtic. Aquests últims models ens han permès explorar les subtileses de diferents elements lingüístics com ara l'impacte que tenen les emocions en la classificació correcta del sentiment expressat en un text. Posteriorment, després d'aquests treballs inicials es van desenvolupar models AP, en particular, Xarxes neuronals convolucionals (XNC) que van ser aplicades a les tasques prèviament esmentades. En el cas de el reconeixement de la personalitat, s'han comparat models clàssics de l'aprenentatge automàtic amb models d'aprenentatge profund la qual cosa a permet establir una comparativa de les dos aproximacions sota les mateixes premisses. Cal remarcar que el PNL ha evolucionat dràsticament en els últims anys gràcies a el desenvolupament de campanyes d'avaluació pública on múltiples equips d'investigació comparen les capacitats dels models que proposen sota les mateixes condicions. La majoria dels models presentats en aquesta tesi van ser o bé avaluats mitjançant campanyes d'avaluació públiques, o bé s'ha emprat la configuració d'una campanya pública prèviament celebrada. Sent conscients, per tant, de la importància d'aquestes campanyes per a l'avanç del PNL, vam desenvolupar una campanya d'avaluació pública on l'objectiu era classificar el tema tractat en un tweet, per a la qual cosa vam recollir i etiquetar un nou conjunt de dades. A mesura que avançàvem en el desenvolupament del treball d'aquesta tesi, vam decidir estudiar en profunditat com les XNC s'apliquen a les tasques de PNL. En aquest sentit, es van explorar dues línies de treball.En primer lloc, vam proposar un mètode d'emplenament semàntic per RNC, que planteja una nova manera de representar el text per resoldre tasques de PNL. I en segon lloc, es va introduir un marc teòric per abordar una de les crítiques més freqüents de l'aprenentatge profund, el qual és la falta de interpretabilitat. Aquest marc cerca visualitzar quins patrons lèxics, si n'hi han, han estat apresos per la xarxa per classificar un text.
[EN] In the last years, Deep Learning (DL) has revolutionised the potential of automatic systems that handle Natural Language Processing (NLP) tasks. We have witnessed a tremendous advance in the performance of these systems. Nowadays, we found embedded systems ubiquitously, determining the intent of the text we write, the sentiment of our tweets or our political views, for citing some examples. In this thesis, we proposed several NLP models for addressing tasks that deal with social media text. Concretely, this work is focused mainly on Sentiment Analysis and Personality Recognition tasks. Sentiment Analysis is one of the leading problems in NLP, consists of determining the polarity of a text, and it is a well-known task where the number of resources and models proposed is vast. In contrast, Personality Recognition is a breakthrough task that aims to determine the users' personality using their writing style, but it is more a niche task with fewer resources designed ad-hoc but with great potential. Despite the fact that the principal focus of this work was on the development of Deep Learning models, we have also proposed models based on linguistic resources and classical Machine Learning models. Moreover, in this more straightforward setup, we have explored the nuances of different language devices, such as the impact of emotions in the correct classification of the sentiment expressed in a text. Afterwards, DL models were developed, particularly Convolutional Neural Networks (CNNs), to address previously described tasks. In the case of Personality Recognition, we explored the two approaches, which allowed us to compare the models under the same circumstances. Noteworthy, NLP has evolved dramatically in the last years through the development of public evaluation campaigns, where multiple research teams compare the performance of their approaches under the same conditions. Most of the models here presented were either assessed in an evaluation task or either used their setup. Recognising the importance of this effort, we curated and developed an evaluation campaign for classifying political tweets. In addition, as we advanced in the development of this work, we decided to study in-depth CNNs applied to NLP tasks. Two lines of work were explored in this regard. Firstly, we proposed a semantic-based padding method for CNNs, which addresses how to represent text more appropriately for solving NLP tasks. Secondly, a theoretical framework was introduced for tackling one of the most frequent critics of Deep Learning: interpretability. This framework seeks to visualise what lexical patterns, if any, the CNN is learning in order to classify a sentence. In summary, the main achievements presented in this thesis are: - The organisation of an evaluation campaign for Topic Classification from texts gathered from social media. - The proposal of several Machine Learning models tackling the Sentiment Analysis task from social media. Besides, a study of the impact of linguistic devices such as figurative language in the task is presented. - The development of a model for inferring the personality of a developer provided the source code that they have written. - The study of Personality Recognition tasks from social media following two different approaches, models based on machine learning algorithms and handcrafted features, and models based on CNNs were proposed and compared both approaches. - The introduction of new semantic-based paddings for optimising how the text was represented in CNNs. - The definition of a theoretical framework to provide interpretable information to what CNNs were learning internally.
Giménez Fayos, MT. (2021). Natural Language Processing using Deep Learning in Social Media [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/172164
TESIS

This thesis aims to perform named entity recognition for English social media texts. Named Entity Recognition (NER) is applied in many NLP tasks as an important preprocessing procedure. Social media texts contain lots of real-time data and therefore serve as a valuable source for information extraction. Nevertheless, NER for social media texts is a rather challenging task due to the noisy context. Traditional approaches to deal with this task use hand-crafted features but prove to be both time-consuming and very task-specific. As a result, they fail to deliver satisfactory performance. The goal of this thesis is to tackle this task by automatically identifying and annotating the named entities with multiple types with the help of neural network methods. In this thesis, we experiment with three different word embeddings and character embedding neural network architectures that combine long short- term memory (LSTM), bidirectional LSTM (BI-LSTM) and conditional random field (CRF) to get the best result. The data and evaluation tool comes from the previous shared tasks on Noisy User-generated Text (W- NUT) in 2017. We achieve the best F1 score 42.44 using BI-LSTM-CRF with character-level representation extracted by a BI-LSTM, and pre-trained word embeddings trained by GloVe. We also find out that the results could be improved with larger training data sets.

Social media users generate a large volume of data, which can contain meaningful and useful information. One such example is information about locations, which may be useful in applications such as marketing and security monitoring. There are two types of locations: location entities mentioned in the text of the messages and the physical locations of users. Extracting the first type of locations is not trivial because the location entities in the text are often ambiguous. In this thesis, we implement a sequential classification model with conditional random fields followed by a rule-based disambiguation model, we apply them to Twitter messages (tweets) and we show that they handle the ambiguous location entities in our dataset reasonably well. Only very few users disclose their physical locations; in order to automatically detect their locations, many approaches have been proposed using various types of information, including the tweets posted by the users. It is not easy to infer the original locations from text data, because text tends to be noisy, particularly in social media. Recently, deep learning techniques have been shown to reduce the error rate of many machine learning tasks, due to their ability to learn meaningful representations of input data. We investigate the potential of building a deep-learning architecture to infer the location of Twitter users based merely on their tweets. We find that stacked denoising auto-encoders are well suited for this task, with results comparable to state-of-the-art models. Finally, we combine the two models above with a third-party sentiment analysis tool and obtain a intelligent social media monitoring system. We show a demo of the system and that it is able to predict and visualize the locations and sentiments contained in a stream of tweets related to mobile phone brands - a typical real world e-business application.

Social media has provided innovative ways for teachers to engage students in the learning process but has created a challenge for teachers to incorporate these applications in a manner that is both meaningful to the learning objectives and acceptable to their administration. While social media in education has been the focus of many studies, research on the implementation of social media within English language arts (ELA) classrooms is limited, leaving ELA teachers and school administrators without a full scope of the educational potential or best practices when using social media for instruction. Accordingly, this study explored the experiences of secondary ELA teachers using social media as an instructional tool. Using a basic qualitative design, this investigation was framed by the concepts of connectivism and convergence culture. The study employed in-depth interviews of 9 teachers chosen through a purposeful sampling of ELA teachers within the United States. Data analysis began with a priori coding of the interview transcripts based on the conceptual framework, followed by a secondary analysis through in vivo coding. The results indicated that social media networks provided teachers with an engaging and relevant approach to connect their content and instruction to students' lives. Results also revealed that while the teachers faced challenges similar to those noted in previous research, these challenges were viewed as opportunities to teach digital literacy within the ELA content rather than as a deterrent. The results of this study may allow teachers to use social media networks as educational tools in alignment with instructional practices to improve student performance.

Social media is a large part of Swedish adolescents’ lives. As such, social media is also prevalent in Swedish upper secondary schools. Some international research has shown positive outcomes from the use of social media as a learning tool in the English as a Foreign Language (EFL) classroom, such as higher student motivation. However, there are no recent equivalent studies in Sweden. Teachers’ attitudes have also been researched, also in a Swedish context. The aim of this study is twofold: to explore EFL teachers’ attitudes, as well as their practices, regarding social media as a learning tool. A web-questionnaire was constructed and subsequently answered by 20 random EFL teachers in Swedish upper secondary schools. The findings show that participants have positive attitudes toward using social media platforms in their teaching. It was also found that participants do indeed use social media, as well as many other tools, in their teaching, and use it for many different purposes. The findings suggest there needs to be more research into how social media platforms could, should, and are being used in Swedish EFL classrooms.

Traditionally, the field of rhetoric and composition has valued long-form essay writing, which requires students to engage patiently and at length with revision. In contrast, students today spend much time outside of school producing fast-paced and short posts for social media. This dissertation argues that students’ social media interactions provide them nuanced, dialogic, and complex rhetorical understandings about writing—but that students need help developing discursive processes to support transfer of their social media knowledge to other writing contexts, including long-form academic writing. Drawing from two semesters of in-class study, I construct for first-year composition classrooms a pedagogy that embraces and cultivates the rhetorical knowledge students gain from social media; I demonstrate how students can analyze, reflect on, and transfer this knowledge to academic contexts. Citing students’ social media and academic writing, I draw from students’ intuitive understandings of the rhetorical concepts medium, context, audience, ethos, and purpose to illustrate how these concepts can productively shift and expand in FYC instruction. To situate this pedagogy within contemporary practices, I analyze leading FYC textbooks and highlight how textbook pedagogies can acknowledge and foreground students’ expanded rhetorical understandings of social media for richer composing processes in all media and for all contexts, digital and non-digital.

An event is something that happens at a time and location. Events include major life events such as graduating college or getting married, and also simple day-to-day activities such as commuting to work or eating lunch. Most work on event extraction detects events and the entities involved in events. For example, cooking events will usually involve a cook, some utensils and appliances, and a final product. In this work, we target the task of determining whether events result in their expected outcomes. Specifically, we target cooking and baking events, and characterize event outcomes into two categories. First, we distinguish whether something edible resulted from the event. Second, if something edible resulted, we distinguish between perfect, partial and alternative outcomes. The main contributions of this thesis are a corpus of 4,000 tweets annotated with event outcome information and experimental results showing that the task can be automated. The corpus includes tweets that have only text as well as tweets that have text and an image.

Computational approaches to opinion mining have mostly focused on polarity detection of product reviews by classifying the given text as positive, negative or neutral. While, there is less effort in the direction of socio-political opinion mining to determine favorability towards given targets of interest, particularly for social media data like news comments and tweets. In this research, we explore the task of automatically determining from the text whether the author of the text is in favor of, against, or neutral towards a proposition or target. The target may be a person, an organization, a government policy, a movement, a product, etc. Moreover, we are interested in detecting the reasons behind authors’ positions. This thesis is organized into three main parts: the first part on Twitter stance detection and interaction of stance and sentiment labels, the second part on detecting stance and the reasons behind it in online news comments, and the third part on multi-target stance classification. One may express favor (or disfavor) towards a target by using positive or negative language. Here, for the first time, we present a dataset of tweets annotated for whether the tweeter is in favor of or against pre-chosen targets, as well as for sentiment. These targets may or may not be referred to in the tweets, and they may or may not be the target of opinion in the tweets. We develop a simple stance detection system that outperforms all 19 teams that participated in a recent shared task competition on the same dataset (SemEval-2016 Task #6). Additionally, access to both stance and sentiment annotations allows us to conduct several experiments to tease out their interactions. Next, we proposed a novel framework for joint learning of stance and reasons behind it. This framework relies on topic modeling. Unlike other machine learning approaches for argument tagging which often require a large set of labeled data, our approach is minimally supervised. The extracted arguments are subsequently employed for stance classification. Furthermore, we create and make available the first dataset of online news comments manually annotated for stance and arguments. Experiments on this dataset demonstrate the benefits of using topic modeling, particularly Non-Negative Matrix Factorization, for argument detection. Previous models for stance classification often treat each target independently, ignoring the potential (sometimes very strong) dependency that could exist among targets. However, in many applications, there exist natural dependencies among targets. In this research, we relieve such independence assumptions in order to jointly model the stance expressed towards multiple targets. We present a new dataset that we built for this task and make it publicly available. Next, we show that an attention-based encoder-decoder framework is very effective for this problem, outperforming several alternatives that jointly learn dependent subjectivity through cascading classification or multi-task learning.

Language students today have grown up with a plethora of technology tools at their fingertips, which has in some cases earned them the title of "digital native". 'Students' high use of technology outside the classroom has led teachers and researchers to believe that technology could be highly effective for language learners when used appropriately. Yet little is known about how students actually react to technology-based tools for language learning purposes. This study seeks to not only understand student attitudes toward technology in general, but also to see how those attitudes might affect student attitudes toward online language learning tools in a social media context. Using a design-based research approach, we implemented a curriculum that utilizes a social networking environment in which students could consume authentic language samples and practice using the language in a controlled environment. Through the analysis of pre and post surveys, it was discovered that age was the most significant predictor of student attitudes toward technology, but that the extent to which students use technology proves to be a more significant predictor when other variables are factored in. Furthermore, it was discovered that general attitudes toward technology do affect the ways in which students will react to a technology-based curriculum. Nevertheless, the way in which a curriculum is presented can be a stronger factor in predicting how the curriculum will be received.

Technology has changed the way we communicate, teach, and learn. Today’s generation of college students has never known what it is like to be without a cell phone or a computer. Social media is an integral part of their lives. As technology evolves, so do students’ expectations for a learning experience. This experience goes beyond the traditional classroom boundaries’ constrains. Understanding how to take advantage of the pedagogical potential of new technologies is therefore essential for language educators. Broadly speaking, there is much discussion about the integration of social media in language education. However, little is known about how these technologies may work in the language classroom. While many studies focus on the pedagogical benefits, few studies have explored the experiences of language educators who use social media to enhance and transform the traditional learning environment. Thus, the purpose of this qualitative exploratory case study was to explore language educators’ perspectives on their use of social media for pedagogical purposes. To gain a better understanding of language educators’ experiences in the use of social media, I interviewed a language educator at a large, public, research university, located in the Southeast of the United States. The primary data sources for this qualitative study consists of a semi-structured survey questionnaire, face-to-face, semi-structured interviews, and a researcher reflective journal. Throughout the study, I analyzed the data to build on the respondents’ comments. To ensure the trustworthiness of the study, I employed several techniques, including data triangulation, peer debriefing, member checking, reflexive journaling, and creating an audit trail. Study findings reveal that participants perceived social media as an instructional medium to blend informal learning into formal learning online, face-to-face, and blended environments, to facilitate a participatory culture, and to provide opportunities for students’ self-expression, self-reflection, and social interaction. Based on the findings, I offer recommendations for instructors, instructional designers, and policy developers. Finally, I address possible future research directions.

What is corporate reputation? How can it be measured? These two questions have been widely discussed by academics, without coming to a shared definition or evaluation methodology. Each research gave its own corporate reputation definition, but all the studies agree on one point: corporate reputation is the result of the relationship between a company and its stakeholders. On the stakeholder’s opinion rely most of the corporate rep- utation measurement techniques that have been proposed during the past years, techniques that were criticized. In this work, we have investigated if corporate reputation can be evaluated from social media data We focused on the Volkswagen scandal and the buzz it created within the Twitter social network. VW’s scandal was chosen because its widely covered evolution through time and its broad effects on VW’s financial performance. In order to fulfill the research goal, tweets about VW (from 28/8/15 - to 6/6/16) were collected. This vast dataset was firstly analyzed and not VW’s related elements were re- moved. The remaining part of it was then classified in two main groups: tweets about VW, but not related to the scandal, and the ones that specifically referred to VW’s wrongdoing. Once the two sets were obtained, each of their elements were evaluated with a sentiment analysis software and after the opinion extraction was calculated the daily aggregated sentiment through a custom-built process, defined to adapt to the Twitter domain. This aggregation produced two different daily sentiment score: the general public opinion about VW and the judgement about the scandal. The work led to excellent results in the not-relevant elements removal phase and the classification one, but the opinion aggregation did not produce significant outcomes. This final results should not be considered as a research drawback, instead they represent a starting point for further analysis on the opinion creation process.

In today’s society there is a large number of social media users that are free to express their opinion on shared platforms. The socio-cultural differences between the people behind those accounts (in terms of ethnicity, gender, sexual orientation, religion, politics, . . . ) give rise to an important percentage of online discussions that make use of offensive language, which often affects in a negative way the psychological well-being of the victims. In order to address the problem, the endless stream of user-generated content engenders a need to find an accurate and scalable solution to detect offensive language using automated methods. This thesis explores different approaches to the offensiveness detection task focusing on five different languages: Arabic, Danish, English, Greek and Turkish. The results obtained using Support Vector Machines (SVM), Convolutional Neural Networks (CNN) and the Bidirectional Encoder Representations from Transformers (BERT) are compared, achieving state-of-the-art results with some of the methods tested. The effect of the embeddings used, the dataset size, the class imbalance percentage and the addition of sentiment features are studied and analysed, as well as the cross-lingual capabilities of pre-trained multilingual models.

Using interpretive qualitative research methods, this dissertation study examines how a telecollaboration in a social networking site affects learners’ intercultural competence development. In particular, it explores the changes in learners’ understanding of the concept of culture, as well as their understanding of their own and target cultures. Additionally, this study investigates what role the medium of the exchange plays in the project. Finally, it analyzes three individual learners’ experiences in an online intercultural exchange. Two intact classes of college-level language learners—U.S. students learning Spanish and Colombian students learning English—participated in an eight-week telecollaboration using Facebook. Each week students shared information about their own culture and reacted to the posts about the target culture made by the students in the partner class. This study analyzed 14 U.S. students’ experiences and learning, based on their pre- and post-telecollaboration questionnaires, reflexive portfolios, and pre- and post-telecollaboration one-on-one interviews. Data analysis demonstrated that the participants experienced a notable development in their intercultural competence. In particular, participants showed a more detailed understanding of the concept of culture, as well as the reasons for cultural learning and awareness. The activity of explaining their own culture to foreigners led the participants to an ongoing analysis of their own culture and, therefore, facilitated development in their understanding of the home culture. Specifically, participants were able to take a critical stance in their analysis and evaluate their own culture from the position of an outsider. Finally, participants gained knowledge about Colombian culture and developed an ability to analyze cultural products and practices, as well as infer information about cultural perspectives from that analysis. In general, the telecollaboration enabled participants to develop critical cultural awareness and change their worldviews from more ethnocentric to more ethnorelative. In regard to the use of Facebook, this popular social networking site proved to be uniquely suitable for a telecollaboration. It gave the participants an opportunity to create a personal learning environment that was accessible whenever and wherever they needed it. It also facilitated a bridging between participants’ learning and leisure online activities. The three case studies demonstrated that while any type of learner can benefit from telecollaboration, their experiences and the degree of intercultural competence development are individual. In addition, the development of intercultural competence appeared to be non-linear, which demonstrates the difficulty of its assessment. Based on the findings, the researcher concludes that telecollaboration facilitates the development of intercultural competence and, thus, is an effective way of addressing the challenge of culture instruction in foreign language courses. Additionally, the unique properties of social networking sites make them an easily accessible platform for telecollaboration projects.

Foreign language pedagogy has often been influenced by findings drawn from the area of first and/or second language acquisition with confusing results. The thesis explores the extent to which the inter-relation of variables in foreign language learning differs from that in natural acquisition processes and varies across a range of learning contexts, thereby encouraging different pedagogical approaches. Chapter one argues that some models underestimate the different variables involved and suggests that a socio-cultural approach is more effective in identiffing and explaining the shifting relation between context and cognition. Chapter two seeks to situate foreign language study within a more general process of jimctional differentiation' in the child's widening linguistic repertoire, arguing that not only does the relation of context and cognition change between L I and L2 but also within L2 itself. Chapter three examines the shifting relation between context and cognition with reference to the language programme of European Schools in general, and the one at Culham in particular. The latter provides a basis for answering two questions: (1) is there a correlation between success in an acquisition poor environment and the extent of the learner's analytic competence; (2) does motivation play an increasing role in affecting success in contexts where goals are long-term rather than immediate? Data collected from the school are analyzed in chapter four. Response to both questions would seem to he positive although the complexity of the learners' backgrounds produced greater variation in the role of affective factors than anticipated. Finally, chapter five argues that the relative success of foreign language study from an early stage in schemes such as the European schools or the immersion programmes depends upon a precise interplay of socio-cognitive variables which is unlikely to he replicated elsewhere.

Word-of-mouth (WOM) communication is a well studied phenomenon in the literature and content propagation in Online Social Networks (OSNs) is one of the forms of WOM mechanism that have been prevalent in recent years specially with the widespread surge of online communities and online social networks. The basic piece of information in most OSNs is a post (e.g., a tweet in Twitter or a post in Facebook). A post can contain different types of content such as text, photo, video, etc, or a mixture of two or more them. There are also various ways to enrich the text by mentioning other users, using hashtags, and adding URLs to external contents. The goal of this study is to investigate what factors contribute into the propagation of messages in Google+. To answer to this question a multidimensional study will be conducted. On one hand this question could be viewed as a natural language processing problem where topic or sentiment of posts cause message dissemination. On the other hand the propagation can be effect of graph properties i.e., popularity of message originators (node degree) or activities of communities. Other aspects of this problem are time, external contents, and external events. All of these factors are studied carefully to find the most highly correlated attribute(s) in the propagation of posts.

The economic burden of mental illness costs Canada billions of dollars every year. Millions of people suffer from mental illness, and only a fraction receives adequate treatment. Identifying people with mental illness requires initiation from those in need, available medical services, and professional experts’ time allocation. These resources might not be available all the time. The common practice is to rely on clinical data, which is generally collected after the illness is developed and reported. Moreover, such clinical data is incomplete and hard to obtain. An alternative data source is conducting surveys through phone calls, interviews, or mail, but this is costly and time-consuming. Social media analysis has brought advances in leveraging population data to understand mental health problems. Thus, analyzing social media posts can be an essential alternative for identifying mental disorders throughout the Canadian population. Big data research of social media may also endorse standard surveillance approaches and provide decision-makers with usable information. More precisely, social media analysis has shown promising results for public health assessment and monitoring. In this research, we explore the task of automatically analysing social media textual data using Natural Language Processing (NLP) and Machine Learning (ML) techniques to detect signs of mental health disorders that need attention, such as depression and suicide ideation. Considering the lack of comprehensive annotated data in this field, we propose a methodology for transfer learning to utilize the information hidden in a training sample and leverage it on a different dataset to choose the best-generalized model to be applied at the population level. We also present evidence that ML models designed to predict suicide ideation using Reddit data can utilize the knowledge they encoded to make predictions on Twitter data, even though the two platforms differ in the purpose, structure, and limitations. In our proposed models, we use feature engineering with supervised machine learning algorithms (such as SVM, LR, RF, XGBoost, and GBDT), and we compare their results with those of deep learning algorithms (such as LSTM, Bi-LSTM, and CNNs). We adopt the CNN model for depression classification that obtained the highest F1-score on the test dataset (0.898) and 0.941 recall. This model is later used to estimate the depression level of the population. For suicide ideation detection, we used the CNN model with pre-trained fastText word embeddings and linguistic features (LIWC). The model achieved an F1-score of 0.936 and a recall of 0.88 to predict suicide ideation at the user-level on the test set. To compare our models’ predictions with official statics, we used 2015-2016 population based Canadian Community Health Survey (CCHS) on Mental Health and Well-being conducted by Statistics Canada. The data is used to estimate depression and suicidality in Canadian provinces and territories. For depression, (n=53,050) respondents filled in the Patient Health Questionnaire-9 (PHQ-9) from 8 provinces/territories. Each survey respondent with a score ≥ 10 on the PHQ-9 was interpreted as having moderate to severe depression because this score is frequently used as a screening cut-point. The weighted percentage of depression prevalence during 2015 for females and males of the age between 15 to 75 was 11.5% and 8.1%, respectively (with 54.2% females and 45.8% males). Our model was applied on a population-representative dataset that contains 24,251 Twitter users who posted 1,735,200 tweets during 2015 with a Pearson correlation of 0.88 for both sex and age within the seven provinces and NT territory included in the CCHS. An age correlation of 0.95 was calculated for age and sex (separately) and our model estimated that 10% of the sample dataset has evidence of depression (58.3% females and 41.7% males). For the second task, suicide ideation, Statistics Canada (2015) estimated the total number of people who reported serious suicidal thoughts as 3,396,700 persons, i.e., 9.514% of the total population, whereas our models estimated 10.6% of the population sample were at risk of suicide ideation (59% females and 41% males). The Pearson correlation coefficients between the actual suicide ideation within the last 12 months and the predicted model for each province per age, sex, and both more than 0.62, which indicates a reasonable correlation.

Thesis: Ph. D., Massachusetts Institute of Technology, School of Architecture and Planning, Program in Media Arts and Sciences, 2015.
Cataloged from PDF version of thesis.
Includes bibliographical references (pages 127-133).
Autism Spectrum Disorder (ASD) is a complex neurodevelopmental disorder that affects approximately 1-in-68 persons and 1-in-42 boys. Challenges related to communication form a core characteristic of the ASD condition. Literature suggests the single most important prognostic indicator for young children diagnosed with ASD is communication ability. Yet, engaging children diagnosed with ASD in communication centered activities is one of the cardinal challenges presented by the condition and contributes to poor outcome. This work explores the development, implementation and testing of StoryScape, a platform for engaging children diagnosed with ASD in communication centered activities. Through extensive work with the ASD community, I have sought to develop a more holistic learning technology around communication. The StoryScape platform connects web, mobile, and physical technologies through interactive stories, enabling fun and engaging learning through story creation and other activities. While the main focus of this work is related to children diagnosed with ASD, I also present a number of exploratory studies conducted with neurotypical children. In total, six exploratory studies, three with children diagnosed with ASD and three with neurotypical children, are presented. These studies demonstrate the many uses of StoryScape, and were critical for eliciting feedback on usability and design features during its iterative development. In addition, I present the results from a seven-week study of story co-creation, conducted as a whole-class activity with students diagnosed with ASD. Results show a statistically significant increase in vocal expressions during use of StoryScape. Students also demonstrate a strong preference for the StoryScape activity compared to other activities. Teachers reported being surprised by the high level of student engagement and cooperation with each other, and they reported that students looked forward to using StoryScape daily. While teachers reported meaningful communication improvements during use of StoryScape for all of their students, one student was described by his teacher as "a totally different student when using StoryScape."
by Micah Rye Eckhardt.
Ph. D.

Overall, this thesis purports to make two significant contributions to knowledge. The first is a foundational critique of political economy in the context of an emergent global knowledge economy. The second is a method for analysing evaluations in language. The relationships that give coherence to those two contributions are as follows. The widely-heralded emergence of a knowledge economy indicates that more intimate aspects of human activity have become exposed to commodification on a massive scale, specifically, activities associated with thought and language. Correspondingly, more abstract forms of value have developed as the products of thought and language have become dominant commodity forms. Historical investigation shows that value has moved from an objective category in political economy, pertaining to such substances as precious metals and land, to become situated today predominantly in “expert” expressions of language, or more precisely, their institutional contexts of production. These are now propagated and circulated on a global scale. Legal, political, and technological developments are key in the development of new, more abstract forms of labour and value, although the relationships connecting these are neither simple nor direct. They are, however, inseparably related in the trajectories that this thesis describes. Consequently they are dealt with inseparably throughout.

Author profiling is a field within Natural Language Processing (NLP) that is concerned with identifying various characteristics and demographic factors of authors, such as gender, age, location, native language, political orientation, and personality by analyzing the style and content of their writings. There is a growing interest in author profiling, with applications in marketing and advertising, opinion mining, personalization, recommendation systems, forensics, security, and defense. In this work, we build several classification models using NLP, Deep Learning, and classical Machine Learning techniques that can identify the gender and age of a Twitter user based on the textual contents of their correspondence (tweets) on the platform. Our SVM gender classifier utilizes a combination of word and character n-grams as features, dimensionality reduction using Latent Semantic Analysis (LSA), and a Support Vector Machine (SVM) classifier with linear kernel. At the PAN 2018 author profiling shared task, this model achieved the highest performance with 82.21%, 82.00%, and 80.90% accuracy on the English, Spanish, and Arabic datasets, respectively. Our age classifier was trained on a dataset of 11,160 Twitter users, using the same approach, though the age classification experiments are preliminary. Our Deep Learning gender classifiers are trained and tested on English datasets. Our feedforward neural network consisting of a word embedding layer, flattening, and two densely-connected layers achieves 79.57% accuracy, and our bidirectional Long Short-Term Memory (LSTM) neural network achieves 76.85% accuracy on the gender classification task.

I am comparing the tactics of Mexican youth movement #Yo Soy 132 and Occupy to better understand how these groups work against the hegemonic views presented by mass media. I aim to understand the media structures in Mexico and the United States through the lens of these social groups and consider how they are similar or different. I also take into consideration societal structures, such as varying levels of Internet access, that influence the way these groups function. These movements are in constant flux and their interaction with the public is changing everyday, but I hope to provide some insight into their tactics and strategies and whether or not they are successful in achieving their established goals.

This thesis examines German philosopher Martin Heidegger’s concepts of being and time, the role of language in being, and ways of authentic being through the lens of modern media practices in the Information Age. It relates Heidegger’s philosophy to the media ecology theory introduced by Marshall McLuhan in the 1960s by exploring McLuhan’s themes of tribal, typographical and electronic man. In addition, this thesis considers the role of mass media in information dissemination. The goal of this report is to explicate the shaping effects of mass media, especially social media, on individual perceptions and societal culture and identify ways in which such shaping affects authentic ways of being.

This work highlights specific lexical items that have become racialized in specific contextual applications and tests how these words are cognitively processed. This work presents the results of a visual world (Huettig et al 2011) eye-tracking study designed to determine the perception and application of racialized (Coates 2011) adjectives. To objectively select the racialized adjectives used, I developed a corpus comprised of popular media sources, designed specifically to suit my research question. I collected publications from digital media sources such as Sports Illustrated, USA Today, and Fortune by scraping articles featuring specific search terms from their websites. This experiment seeks to aid in the demarcation of socially salient groups whose application of racialized adjectives to racialized images is near instantaneous, or at least less questioned. As we view growing social movements which revolve around the significant marks unconscious assumptions leave on American society, revealing how and where these lexical assignments arise and thrive allows us to interrogate the forces which build and reify such biases. Future research should attempt to address the harmful semiotics these lexical choices sustain.

In current study, a mechanism to extract traffic related information such as congestion and incidents from textual data from the internet is proposed. The current source of data is Twitter, however, the same mechanism can be extended to any kind of text available on the internet. As the data being considered is extremely large in size automated models are developed to stream, download, and mine the data in real-time. Furthermore, if any tweet has traffic related information then the models should be able to infer and extract this data. To pursue this task, Artificial Intelligence, Machine Learning, and Natural Language Processing techniques are used. These models are designed in such a way that they are able to detect the traffic congestion and traffic incidents from the Twitter stream at any location. Currently, the data is collected only for United States. The data is collected for 85 days (50 complete and 35 partial) randomly sampled over the span of five months (September, 2014 to February, 2015) and a total of 120,000 geo-tagged traffic related tweets are extracted, while six million geo-tagged non-traffic related tweets are retrieved. The classification models for detection of traffic congestion and incidents are trained on this dataset. Furthermore, this data is also used for various kinds of spatial and temporal analysis. A mechanism to calculate level of traffic congestion, safety, and traffic perception for cities in U.S. is proposed. Traffic congestion and safety rankings for the various urban areas are obtained and then they are statistically validated with existing widely adopted rankings. Traffic perception depicts the attitude and perception of people towards the traffic. It is also seen that traffic related data when visualized spatially and temporally provides the same pattern as the actual traffic flows for various urban areas. When visualized at the city level, it is clearly visible that the flow of tweets is similar to flow of vehicles and that the traffic related tweets are representative of traffic within the cities. With all the findings in current study, it is shown that significant amount of traffic related information can be extracted from Twitter and other sources on internet. Furthermore, Twitter and these data sources are freely available and are not bound by spatial and temporal limitations. That is, wherever there is a user there is a potential for data.

This study is an exploratory look into the use of text-based slang, or textisms, in modern communication. People use textisms in multiple media of communication, whether it is in text messages, emails, and various types of social media. This study asked a focus group about their uses and gratifications from textisms. The focus group delved into their appreciations as both users and receivers. Participants reflected on their personal experience and preferences with textisms for their personal communications as well as their opinion for various types of businesses using them in their larger message design efforts. The participants were comprised of youthful twenty to twenty-two year olds in undergraduate studies at the University of South Florida. Each was currently enrolled in a media studies course at the time of the discussion group. Each member of the nine person group provided information. Each was asked as an individual to reflect on specific questions while also adding their own opinion to the larger group discussion. General themes pulled from the discussion was the creation of an author's personality, textisms use as a way to create tone, and the varying levels of context needed within personal communication. Participants were in agreement that an individual could be characterized by his or her message design, especially through that individual's use of textisms. By establishing the author's personality, textisms created expectations from the audience. They help reinforce the relationship of those engaged. Tone is often removed from simple text-based communication. Through the use of textisms, or lack of, tone is able to be added to short messages with the use of a single textism. Playfulness, sarcasm, and seriousness are all able to be quickly established by an author with the additive of a textism. Context is always something to be aware of when interacting with any audience. Gender, generational, formality, and overall relationship are all aspects that impact the interpretation of a message, especially a message that contains textisms. Being aware of how a gender, an age group, or an individual will receive a message further add to the relevance of that message. More information is always best, but if there are limitations, then there are general approaches that can be used to help develop the most salient message. Businesses had a specific set of expectations from the discussion group that shaped their responses. The consensus for this group was that textisms should be refrained from use within any business context. The group did allow for leeway when it came to more casual style businesses ranging from fast-food restaurants to youth-targeted products, but ultimately said there could be a confusion if a company with an older demographic was using textisms, as it would be appear to be targeting a younger audience. As this group is comprised of young people with a vision for their futures within the business world, their approach to textisms in business was also attached to that vision. They see business as very formal, and that formality is expected to translate into all areas, including communications. This study pulled valuable overall themes for an exploratory study. The specific details of a small sample group could and should be tested for greater validity, such as the strong and opposing reaction to varying types of smiley faces. The group had conflicting expectations for certain types of textisms and these expectations were from a narrow set of demographics. Further inquiry into this topic will lead to a better understanding of how language is evolving and how those changes are being used in today's and tomorrow's communication.

[EN] The possibility of knowing people traits on the basis of what they write is a field of growing interest named author profiling. To infer a user's gender, age, native language or personality traits, simply by analysing her texts, opens a wide range of possibilities from the point of view of forensics, security and marketing. Furthermore, social media proliferation, which allows for new communication models and human relations, strengthens this wide range of possibilities to bounds never seen before. Idiosyncrasy inherent to social media makes them a special environment of communication, where freedom of expression, informality and spontaneous generation of topics and trends, enhances the knowledge of the daily reality of people in their use of language. However, the same idiosyncrasy makes difficult, or extremely costly, the application of linguistic techniques. In this work we have proposed EmoGraph, a graph-based approach with the aim at modelling the way that users express their emotions, and the way they include them in their discourse, bearing in mind not only their frequency of occurrence, but also their position and relationship with other elements in the discourse. Our starting hypothesis is that users express themselves and their emotions differently depending on their age and gender, and besides, we think that this is independent on their language and social media where they write. We have collaborated in the creation of a common framework of evaluation at the PAN Lab of CLEF, generating resources that allowed us to verify our hypothesis achieving comparable and competitive results with the best ones obtained by other researchers on the field. In addition, we have investigated whether the expression of emotions would help to differentiate among users of different varieties of the same language, for example, Spanish from Spain, Mexican and Argentinian, or Portuguese from Portugal and Brazil. Our hypothesis is that the variation among languages is based more on lexical aspects, and we have corroborated it after comparing EmoGraph with representations based on word patterns, distributed representations and a representation that uses the whole vocabulary, but reducing its dimensionality to only 6 features per class, what is suitable for its application to big data environments such as social media.
[ES] La posibilidad de conocer rasgos de una persona a partir únicamente de los textos que escribe se ha convertido en un área de gran interés denominada author profiling. Ser capaz de inferir de un usuario su sexo, edad, idioma nativo o los rasgos de su personalidad, simplemente analizando sus textos, abre todo un abanico de posibilidades desde el punto de vista forense, de la seguridad o del marketing. Además, la proliferación de los medios sociales, que favorece nuevos modelos de comunicación y relación humana, potencia este abanico de posibilidades hasta cotas nunca antes vistas. La idiosincrasia inherente a estos medios sociales hace de ellos un entorno de comunicación especial, donde la libertad de expresión, la informalidad y la generación espontánea de temáticas y tendencias propician el acercamiento a la realidad diaria de las personas en su uso de la lengua. Sin embargo, esa misma idiosincrasia hace que en muchas ocasiones la aplicación de técnicas lingüísticas de análisis no sea posible, o sea extremadamente costoso. En este trabajo hemos propuesto EmoGraph, una representación basada en grafos con el objetivo de modelar el modo en que los usuarios expresan sus emociones, y el modo en que las articulan en el marco de su discurso, teniendo en consideración no sólo su frecuencia, sino también su posición y relación con y respecto a los elementos del mismo. Nuestra hipótesis de partida es que los usuarios se expresan y expresan sus emociones de manera diferente dependiendo de su edad y sexo, y además, pensamos que esto es así independientemente de su idioma y del medio donde escriban. Hemos colaborado en la creación de un marco común de evaluación en el laboratorio PAN del CLEF, generando recursos que nos han permitido verificar nuestra hipótesis y conseguir resultados comparables y competitivos con los mejores resultados obtenidos por los investigadores del área. Además, hemos querido investigar si la expresión de emociones permitiría diferenciar entre hablantes de diferentes variedades de una misma lengua, por ejemplo españoles, mexicanos o argentinos, o portugueses y brasileños. Nuestra hipótesis es que la variación entre lenguas se basa más en aspectos léxicos, y así lo hemos corroborado tras comparar EmoGraph con representaciones basadas en patrones, representaciones distribuidas y una representación que toma en consideración el vocabulario completo, pero reduciendo su dimensionalidad a únicamente 6 características por clase y que se erige idónea para su aplicación en entornos big data como los medios sociales.
[CAT] La possibilitat de conèixer trets d'una persona únicament a partir dels textos que escriu s'ha convertit en una àrea de gran interès anomenada author profiling. Ser capaç d'inferir d'un usuari el sexe, l'edat, l'idioma nadiu o els trets de la seua personalitat tan sols analitzant els seus textos, obre tot un ventall de possibilitats des del punt de vista forense, de la seguretat o del màrketing. A més, la proliferació dels mitjans socials, que afavoreix nous models de comunicació i de relació humana, potencia aquest ventall de possibilitats fins a cotes que no s'han vist fins ara. La idiosincràsia inherent a aquests mitjans socials en fa d'ells un entorn de comunicació especial, on la llibertat d'expressió, la informalitat i la generació espontània de temàtiques i tendències propicien l'aproximació a la realitat diària de les persones en l'ús que fan de la llengua. Tanmateix, aquesta idiosincràsia fa que en moltes ocasions no es puguin aplicar tècniques lingüístiques d'anàlisi, o que fer-ho resulti extremadament costós. En aquest treball hem proposat EmoGraph, una representació basada en grafs que té l'objectiu de modelar la manera en què els usaris expressen les seves emocions, i la manera com les articulen en el marc de llur discurs, considerant-ne no només la freqüència sinó també la posició i la relació amb i respecte als elements del discurs. La nostra hipòtesi de partida és que els usuaris s'expressen i expressen llurs emocions de manera diferent depenent de l'edat i el sexe, i a més, pensem que això és així independentment de l'idioma i del mitjà en què escriguin. Hem col·laborat en la creació d'un marc comú d'avaluació al laboratori PAN del CLEF, generant recursos que ens han permès verificar la nostra hipòtesi i aconseguir resultats comparables i competitius amb els millors resultats obtinguts pels investigadors de l'àrea. A més, hem volgut investigar si l'expressió d'emocions permetria establir diferències enre parlants de diferents varietats d'una mateixa llengua, per exemple espanyols, mexicans o argentins, o portuguesos i brasilers. La nostra hipòtesi és que la variació entre llengües es basa més en aspectes lèxics, i així ho hem corroborat després de comparar EmoGraph amb representacions basades en patrons, representacions distribuïdes i una representació que considera el vocabulari complet, però reduint-ne la dimensionalitat només a 6 característiques per classe i que s'erigeix de manera idònia per a aplicar-la en entorns big data com els mitjans socials.
Rangel Pardo, FM. (2016). Author Profiling en Social Media: Identificación de Edad, Sexo y Variedad del Lenguaje [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/67270
TESIS

This essay describes a case study of textual communication on the online social mediaplatform Tumblr. The aim of the research was to analyse the communicative functions and thestylistic, sociolinguistic, and pragmatic characteristics of textual discourse on Tumblr. Usingdata gathered on Tumblr, this essay analyses discursive aspects such as function, style, andpragmatics, relating the findings to literature on topics including relevance theory and theoriesof computer-mediated communication and cyberpragmatics. The research found thatdiscourse on Tumblr is largely phatic in nature, with an overwhelming focus on the discussionof shared interests. Such discussion on Tumblr appears to heavily depend on sharedbackground knowledge, which works as a barrier of in-group discursive solidarity andprovides and furthers feelings of connectedness. This study further found that alternative useof style and grammar on Tumblr appears to be largely uniform, suggesting that divergencesfrom standard norms are a way of displaying membership of a group and obtaining covertprestige, rather than a display of linguistic innovation.

This thesis is focused upon the development of the computer as a communication medium in history. To accomplish this, the computer is understood as constructed by language and technology where these are in turn grounded upon their roles as forms of cultural mediation. As methodology is of paramount importance, the digital medium is heuristically employed to discuss the epistemic and phenomenological significance of communications media. The more general inference therefore concerns the role of socially constructed media in the fabric of cultural development. In addressing this concern, the paper finds that communications media stand as the repositories of knowledge in the form of artificial memory and figurative technique. The computer then, as a medium in history, may arguably be declared as a paradigmatic instantiation of this role.

The availability of Web 2.0 tools and multiple modalities through digital media is promoting a growing renaissance in linguistic diversity and cultural affiliations, providing a cosmopolitan and plurilingual and multicultural landscape for multilingual users. Full participation in these digitally-mediated activities involves not only print-based literacy but also new literacies that are emerging within Internet-mediated social and communicative contexts. In an effort to better understand how these communication technologies can be used to enhance second language acquisition (SLA), this study explores the relationship between social networking and second language (L2) learning. Grounded within theoretical frameworks of an ecological approach to language (van Lier, 2004), second language socialization (Duff, 2008), and new literacies (Lankshear & Knobel, 2006), this dissertation examines use of social networking sites (SNS) by L2 learners/users of English as a group and as individuals over time in social networking communities through a mixed method approach, including quantitative (e.g., survey) and qualitative (e.g., case study) methods. The ultimate goal is not simply to describe the SNS use by L2 users, but to apply the findings to L2 writing pedagogy that can bridge students' in-school and out-of-school literacy practices and to examine the efficacy of that pedagogy. The three interrelated studies are comprised of 1) a survey-based study of SNS literacy practices and L2 learning, 2) a longitudinal case study of two L2 users' SNS-mediated community investment and identity formation, and 3) a study of the efficacy of an SNS-enhanced genre-awareness instructional unit in an ESL writing classroom. Findings show that L2 users, across culturally diverse groups, performed quantitatively and qualitatively differently in social media usage and displayed different culturally-informed patterns of technological affordances. The longitudinal case study on two users shows that the availability of Web 2.0-mediated semiotic resources allows users to perform complex identity work and explore multimodal selves over time. Implications are that L2 users can gain access to, develop new identities in, and acquire social capital in new communities. Results from the pedagogical intervention show that writing instructions using an SNS-enhanced genre-awareness approach can develop critical awareness of genres across both traditional and digital media.

This dissertation tracks the social circulation of media scripts and collaborative meaning-making in urban Moroccan and Lebanese families' domestic conversations as ways in which the social imaginary of a differentiated pan-Arab audience imaginary is performed. Media scripts refer to television input or information circulated through entextualization processes, embedded direct and indirect quotations framed by a particular discussion, in household dialogues. They include stories, statistics, historical dates, anecdotal observations, music tunes, quotes, iconic units of language varieties and their attendant identities that Moroccan and Lebanese families managed in interpretive discussions. Scripts are easily detached and mobile sound bites that serve on an affective level as possible identity performances. I argue that Fassi Moroccan and Beiruti families are interpretive communities created and who participate in creating a culture of circulation, which is not just about the objects moving through a culture, but the means, methods, and mechanisms of transmission and interpretation built around and negotiated by the members of that community (Lee and LiPuma 2002). Collaborative in this dissertation draws on the Bakhtinian concept that all interaction involves interlocutors, whether present or not, and a set of interpretive conditions affecting meaning (Bakhtin and Holquist 1982:424). Although the social imaginary of an Arab audience is perceived as unitary enough to merit regional satellite programming, the performances of Moroccan and Lebanese families illuminate the differentiated and fractured construction of a pan-Arab cultural project. Through domestic media ethnography of pan-Arab and national entertainment, talk shows, and news programming reception, I explore functional literacies tied to intervisual cues and the management of intergenerational authority; a pan-Arab language ideology that includes performances of multilingualism and shifting identity alignments linked to specific features of linguistic varieties encountered via television; and the link between language, gender, and confessionalism in morality evaluations of gendered media representations. I focus on the everyday domestic contexts, linguistic mechanisms, and discursive frameworks activated by Moroccan and Lebanese families in media engagements.

This research examines the confluence of consumers’ use of social media to share information with the ever-present need for innovative research that yields insight into consumers’ economic decisions. Social media networks have become ubiquitous in the new millennium. These networks, including, among others: Facebook, Twitter, Blog, and Reddit, are brimming with conversations on an expansive array of topics between people, private and public organizations, governments and global institutions. Preliminary findings from initial research confirms the existence of online conversations and posts related to matters of personal finance and consumers’ economic outlook. Meanwhile, the Consumer Confidence Index (CCI) continues to make headline news. The issue of consumer confidence (or sentiment) in anticipating future economic activity generates significant interest from major players in the news media industry, who scrutinize its every detail and report its implications for key players in the economy. Though the CCI originated in the United States in 1946, variants of the survey are now used to track and measure consumer confidence in nations worldwide. In light of the fact that the CCI is a quantified representation of consumer sentiments, it is possible that the level of confidence consumers have in the economy could be deduced by tracking the sentiments or opinions they express in social media posts. Systematic study of these posts could then be transformed into insights that could improve the accuracy of an index like the CCI. Herein lies the focus of the current research—to analyze the attributes of data from social media posts, in order to assess their capacity to generate insights that are novel and/or complementary to traditional CCI methods. The link between data gained from social media and the survey-based CCI is perhaps not an obvious one. But our research will use a data extraction tool called NetBase Insight Workbench to mine data from the social media networks and then apply natural language processing to analyze the social media content. Also, KH Coder software will be used to perform a set of statistical analyses on samples of social media posts to examine the co-occurrence and clustering of words. The findings will be used to expose the strengths and weaknesses of the data and to assess the validity and cohesion of the NetBase data extraction tool and its suitability for future research. In conclusion, our research findings support the analysis of opinions expressed in social media posts as a complement to traditional survey-based CCI approaches. Our findings also identified a key weakness with regards to the degree of ‘noisiness’ of the data. Although this could be attributed to the ‘modeling’ error of the data mining tool, there is room for improvement in the area of association—of discerning the context and intention of posts in online conversations.

This quasi-experimental study examined the impact of interactions with native French language Facebook posts on beginning French language learners' willingness to communicate (WTC) and their attitudes towards the target language and culture in a university setting. In addition, the degree of interaction, by participants, with the French language Facebook posts was recorded and analyzed. This study was conducted during the Spring 2013 semester at the University of Central Florida in Orlando, Florida. Participants in this study were recruited from two sections of FRE 1120, Elementary French Language and Civilization I. Native French language Facebook posts were “pushed” to participants' personal Facebook News Feeds over the course of four weeks, with posts pushed on weekdays only and Facebook polls asking for participant feedback on Fridays. Two instruments were used in this study to obtain participants' demographic information and to measure willingness to communicate as well attitudes towards the target language and culture. In addition, the researcher gathered observational data directly from Facebook. Data were analyzed using a Split-plot ANOVA and descriptive statistics. A total of 26 participants completed the study, with 14 participants in the control group and 12 participants in the treatment group. Both sections of FRE 1120 were conducted in a traditional, face-to-face format and were taught by the same instructor. Results indicated that participants' willingness to communicate in French and their attitudes towards the target language and culture were not significantly impacted by interaction with native French language Facebook posts. The level of Facebook-facilitated interactions in all areas, including “Liking,” Sharing,” and “Commenting” was low. Self-reported interactions, including reading, viewing and translating of French language Facebook posts; Reading and viewing posts (such as simply viewing a photo) was the most frequently reported interaction, with “Commenting” and “Sharing” was the least common interaction. Opportunities for future research are numerous and include increasing the size of the sample, increasing the length of the study, and selected participants' who are more advanced in their mastery of the target language. The potential of social network sites to serve as digitally immersive environments for foreign language learners should be explored in more depth and across various languages.
Ph.D.
Doctorate
Dean's Office, Education
Education and Human Performance
Education; Instructional Technology

This study investigates how non-profit organizations market their organizations and persuades their audience on two different social media platforms. The aim is to examine four non-profit organizations’ use of persuasive language in their Facebook and Twitter posts.  in their Facebook and Twitter posts on their social media pages. The aim is to understand if, how and what linguistic means was are used in the posts through Aristotle’s triangle of rhetoric. His theory consists of the three appeals: ethos, logos and pathos. These three appeals all holds a number of individual features.  with a perspective from classic rhetoric, namely Aristotle’s triangle of rhetoric. The persuasion was examined using Aristotle’s triangle of rhetoric and its appeals ethos, logos, and pathos. The research was carried out by classifying the posts by the three appeals and their features. These defined what linguistic means were used and how. All three appeals are equally important when persuading an audience  public through a text as they are necessary to raise an understanding and interest to the text’s focus. Aristotle argued that when including all three appeals, the text has reached its full extent potential of persuasion. The appeals have their unique attributes and may also persuade when used individually. The An author’s credibility and trust applies to the appeal ethos, logic and reasoning in a of the text applies to logos, and a text that moves its reader’s emotions applies to pathos. The research resulted that aWhether a text’s length matters for its persuasion through a comparison of  the two social media platformshas also been examined through comparing the two social media platformsw. Non-profit organizations do not strive for any profit which make it a challenge for them how to market persuade their organization audience in order and actually be able to be able to continue their work. Many of the posts researched included the appeal pathos which aim mainly to evoke emotions with the readers.

Digital media has a huge impact on today‟s society. Technologies give opportunities for teachers to access different tasks, exercises, games and videos for teaching students a new language. This also means that teachers have opportunity to conduct more varied lessons to motivate students‟ learning. However, digital media comes with advantages as well as disadvantages. This project will therefore discuss the use of digital media for learning English as a foreign language and discover what teachers thinks about using digital media for teaching. To examine the use of digital media in today‟s school and teachers‟ views of using digital media for learning. This project will carry out interviews and observations on teachers from a selected school. Participants from the selected school are teachers who use digital media regularly to teach Swedish students the English language. Moreover, the participants also discussed the advantages and disadvantages with using of digital media in teaching.

Thesis (MA)--University of Stellenbosch, 2010.
ENGLISH ABSTRACT: The aim of this study is to investigate Afrikaner identity in the Afrikaans blogosphere on the basis of the Afrikaner nasionalist identity that was forged during the twentieth century. The environment within which blogs function and its characteristics are discussed, after which the circumstances that led to the development of Afrikaner identity as well as the current sociopolitical position of Afrikaners are traced. Through quantitative analysis, the Afrikaans blogosphere is compared to this identity to discover too what extent the identity is still remnant in Afrikaans bloggers. It is concluded that Afrikaner nationalism is virtually extinct in the Afrikaans blogosphere and that the plurality of identities hosted by it are fragmented and paradoxical.
AFRIKAANSE OPSOMMING: Die doel van hierdie studie is om identiteitsvorming in die Afrikaanse blogosfeer te ondersoek aan die hand van die Afrikaner-nasionalistiese identiteitskonstruk van die twintigste eeu. Aandag word geskenk aan die milieu waarbinne blogs funksioneer en hul aard, waarna die omstandighede waaronder Afrikaneridentiteit ontstaan sowel as die huidige sosio-politiese posisie van Afrikaners nagespeur word. Aan die hand van kwantitatiewe ondersoekmetodes word Afrikaanse blogs dan met die voorafgenoemde identiteit vergelyk in 'n poging om agter te kom in watter mate dit nog by Afrikaanse bloggers teenwoordig is. Die gevolgtrekking word gemaak dat Afrikaner-nasionalisme bykans afwesig is in die Afrikaanse blogosfeer en dat die pluraliteit van identiteite daarop gefragmenteerd en paradoksaal daar uitsien.

In this dissertation we propose algorithms for the analysis of social media texts, focusing on two particular aspects: irony and emojis. We propose novel automatic systems, based on machine learning methods, able to recognize and interpret these two phenomena. We also explore the problem of topic bias in sentiment analysis and irony detection, showing that traditional word based systems are not robust when they have to recognize irony on a new domain. We argue that our proposal is better suited for topic changes. We then use our approach to recognize another phenomena related to irony: satirical news in Twitter. By relying on distributional semantic models, we also introduce a novel method for the study of the meaning and use of emojis in social media texts. Moreover, we also propose an emoji prediction task that consists in predicting the emoji present in a text message using only the text. We have shown that this emoji prediction task can be performed by deep-learning systems with good accuracy, and that this accuracy can be improved by using images included in the post.
En esta tesis proponemos algoritmos para el análisis de textos de redes sociales, enfocándonos en dos aspectos particulares: el reconocimiento automático de la ironía y el análisis y predicción de emojis. Proponemos sistemas automáticos, basados en métodos de aprendizaje automático, capaces de reconocer e interpretar estos dos fenómenos. También exploramos el problema del sesgo en análisis del sentimiento y en la detección de la ironía, mostrando que los sistemas tradicionales, basados en palabras, no son robustos cuando los datos de entrenamiento y test pertenecen a dominios diferentes. El modelo que se propone en esta tesis para el reconocimiento de la ironía es más estable a los cambios de dominio que los sistemas basados en palabras. En una serie de experimentos demostramos que nuestro modelo es también capaz de distinguir entre noticias satíricas y no satíricas. Asimismo, exploramos con modelos semánticos distribucional, si y cómo el significado y el uso de emojis varía entre los idiomas, así como a través de las épocas del año. También nos preguntamos si es posible predecir el emoji que un mensaje contiene solo utilizando el texto del mensaje. Hemos demostrado que nuestro sistema basado en deep-learning es capaz de realizar esta tarea con buena precisión y que se pueden mejorar los resultados si además del texto se utiliza información sobre las imágenes que acompañan al texto.

Via a systematic literature review, this thesis investigates the possible benefits of employing social media as a learning tool in English as a second language classroom. The thesis is focused on the motivation for learning, potential grammar and vocabulary gains as results of using social media at the upper-secondary level. This systematic literature review is relevant since the current Swedish upper-secondary school’s curriculum promotes technology implementation and at the same time social media has become an everyday feature of many students’ lives. In order to measure the possible benefits of social media, this thesis uses a modified version of Koole’s (2009, p. 27) The Framework for the Rational Analysis of Mobile Education (FRAME) model, which originally was used to determine the optimal way for learning on a mobile platform. This thesis adapted that model and altered it to be applicable for social media platforms instead of mobile devices. The literature review found relevant peer-reviewed studies for this topic through a scholarly education database. The age of the participants in the chosen studies vary from 15-19 years old, and all the chosen studies focus on English as a second language in relation to an education system. The reviewed studies show that social media usage is mainly beneficial for motivation to enhance English vocabulary and grammar proficiency. However, social media usage does not show any significant benefits for grammar proficiency and vocabulary gains when students engage with social media. Although motivation to partake on various social media platforms in ESL does show that students tend to have a desire to expand their ESL knowledge. This review study is in agreement with the reviewed studies regarding the potential benefits of social media for ESL. The reviewed studies on their own were inconclusive as to why social media usage was beneficial. By comparing all the studies result, this literature review thesis promotes the idea that it is the social factor from engaging with social media that is the main reason that motivates students to acquire further ESL knowledge in a school setting or extramurally. Thus, the thesis suggests that an integration of social media in the Swedish upper-secondary school should be taken into consideration due to its proven motivational benefits

In this dissertation, we analyze different aspects of the language used in texts published along different social media, and we propose a set of methods for the automatic extraction of polar adjectives as well as for the automatic classification of these texts. First of all, we propose a new classification of polar adjectives according to their lexical features, based on a case study. Secondly, we implement a new domain adaptable system for the automatic extraction of polar adjectives (along with their polarity values), reducing the use of external language resources. Finally, we propose two automatic classifiers (one rule-based and one based on Decision Trees) to identify documents belonging to different stages of the purchase process and texts that analyze different aspects of the product.
En esta tesis, analizamos diferentes aspectos del lenguaje utilizado en los textos publicados en diferentes medios sociales y proponemos una serie de métodos para la extracción automática de adjetivos de opinión, así como para la clasificación automática de dichos textos. En primer lugar, proponemos una nueva clasificación de los adjetivos de opinión de acuerdo con sus características léxicas, basada en un estudio de caso. En segundo lugar, implementamos un nuevo sistema de extracción automática de adjetivos de opinión (junto con sus valores de polaridad), adaptable al dominio y que reduce el uso de recursos lingüísticos externos. Finalmente, proponemos dos clasificadores automáticos (uno basado en reglas y otros basados en ´arboles de decisión) para identificar textos pertenecientes a distintas fases del proceso de compra y textos que analizan diferentes aspectos del producto.
En aquesta tesi, analitzem diferents aspectes del llenguatge utilitzat en els textos publicats en diferents mitjans socials i proposem una sèrie de mètodes per a l’extracció automàtica d’adjectius d’opinió així com per a la classificació automàtica d’aquests textos. En primer lloc, proposem una nova classificació dels adjectius d’opinió, basada en un estudi de cas, més d’acord amb les seves característiques lèxiques. En segon lloc, vam implementar un nou sistema d’extracció automàtica d’adjectius d’opinió (juntament amb els seus valors de polaritat), adaptable al domini i que redueix l’´us de recursos lingüístics externs. Finalment, proposem dos classificadors automàtics (un basat en regles i un altre basats en arbres de decisió) per identificar textos que pertanyen a diferents fases del procés de compra i textos que analitzen diferents aspectes del producte.

This research employed a case study approach to understand emerging themes that may be garnered through documenting the lived experiences of online Deaf activists who have used the video feature available through social media outlets, such as YouTube, as a way to overcome the language barrier typically present for linguistic minorities who are leading social movements within an English-speaking, hearing majority. The focus of this study was the members of the Deaf Community that have taken to an online podium in their fight for autonomy and equality. They champion their Deaf identity, their right to agency and autonomy in areas of language, access, education and employment, in what has exploded into the largest social movement in their cultural history. Therefore, two questions were at the center of this research: 1. "How has experiencing audism affected the lives of Deaf people?", and 2. "How has the use of social media as a platform to fight against audism through natural linguistic expression in American Sign Language impacted that experience?". The growth of individual Deaf identity has created a community action network for the Deaf Community, and access to the technology of videophones and instant access to wireless Internet has brought with it the use of video blogs, or vlogs, within the Deaf Community at explosive rates. The movement from disability to a place of diversity and cultural, ethnic and linguistic minority personhood for the Deaf is a path that is still being forged. Presented in this study is a glimpse into this journey, through a case study of their lived experience.

Thesis: S.M., Massachusetts Institute of Technology, School of Architecture and Planning, Program in Media Arts and Sciences, 2018.
Cataloged from PDF version of thesis.
Includes bibliographical references (pages 75-84).
Learning language and literacy at a young age is important, as children's early language ability can impact their later educational success [1][2]. However, one of the major barriers to early language and literacy learning for many children around the globe is a lack of resources in homes and schools. A variety of technological interventions, such as TV series and educational apps, were designed to help overcome such barriers and support children's learning. However, not all of them necessarily provide children with conversational experiences, which have been found to significantly impact the children's language-related neural development [3]. Among a variety of educational media, embodied interactive agents (e.g., social robots) seem to be an effective yet resource-efficient tool that can enable children to learn through conversational turn taking. Specifically, embodied interactive agents can serve as learning companions for young children and provide more interactive and immersive learning experience. I explored how social robots could help promote children's language and literacy learning. More specifically, I designed and computationally created a collaborative, engaging learning interaction between a robot and a child who play as peers. First, I designed a tablet-based literacy learning game called WordQuest using the design principles for educational games. Second, I developed a reinforcement learning model that enabled the robot to adaptively switch its collaborative roles (e.g., expert and novice roles) in a way that promoted children's best learning. Third, I conducted an experiment with three conditions, which were fixed expert robot, fixed novice robot, and adaptive role switching robot, and tested on 60 children recruited from a local primary school in Boston. Last, I evaluated how the robot's collaborative roles differentially affected children's learning performance, engagement, and perception of the learning experiences. I found out that children across the three conditions all learned new words and had a very positive experience of playing WordQuest with the robot. In addition, children interacting with the adaptive robot consistently outperformed children from the other two conditions in terms of vocabulary acquisition and retention.
by Huili Chen.
S.M.

In  the  society  of  today  social  media  occupies  an  ever  increasingly  part  in  the  life  of  the  modern  man.  Social  media  has  become  a  common  tool  in  communicating  thoughts,  ideas  and  information  for  companies,  authorities  and  private  individu--‐  als  alike.  This papers purpose was that through a case study examine what form an authority’s textual communication through social media (in the form of Facebook and Twitter), web page and traditional ads takes and also to examine possible differences. The case study used a present time perspective and was conducted on Försvarsmakten. A modified manual text analysis was conducted on chosen objects from the four previously mentioned media. The manual text analysis was taken from a given template and was then modified through a frame of reference that included theory about text, rhetoric, genre, text type and advertising. The analyses show that there are both similarities and differences in the com--‐ munication through the different types of media. The purpose of the ads is to dif--‐ ferentiate the image of Försvarsmakten, to inform about Försvarsmakten and also to recruit new soldiers to Försvarsmakten. The purpose of the communica--‐ tion through the web page and the social media is not as distinct, and is shaped by topical events. Försvarsmakten uses a leadership tonality throughout its communication but tends to get more personal and informal through Facebook and Twitter. The pursued argumentative method is always closed and Forsvars--‐ makten does not seem to have any interest in discussing its activities through Facebook or Twitter in the analysed communication. The communication through social media arises in certain cases from a private persons question to Försvarsmakten. Although to claim that there’s a shift in the balance of power in the communication through social media because of that is not fair. On what terms information is spread and discussed is still mainly Försvarsmaktens. Altough it is fair to claim that the Facebook page has become a new meeting place for those who take an interest in Försvarsmakten and their activities. Keywords: textual communication, social media, case study, authority, manual text.
I dagens samhälle tar de sociala medierna allt mer utrymme i den moderna man--‐ niskans verklighet. De sociala medierna har blivit ett vanligt verktyg för att kommunicera tankar, idéer och information för såväl företag som myndigheter och privatpersoner. Det här arbetets syfte var att genom en fallstudie undersöka hur en myndighets textuella kommunikation genom sociala medier (i form av Facebook och Twitter) samt hemsida och traditionella annonser ser ut, samt att undersöka eventuella skillnader. Fallstudien har haft ett nulägesperspektiv och har utförts på For--‐ svarsmakten. Utifrån utvalda objekt från de fyra nämnda medierna gjordes en anpassad brukstextanalys. Brukstextanalysen togs fram utifrån en given mall men anpassades utifrån en referensram där teori kring bland annat text, retorik, genre, texttyp och reklam ingick. Analyserna visar att det finns både skillnader och likheter i kommunikationen mellan de olika medierna. Annonsernas syfte är att nyansera bilden av For--‐ svarsmakten, informera om Försvarsmakten samt att rekrytera nya soldater till Försvarsmakten. Syftet med kommunikationen via hemsidan samt de sociala medierna är inte lika tydligt, och formas efter aktuella skeenden. Forsvarsmak--‐ ten använder sig genomgående av en ledarskapstonalitet men tenderar att bli personliga och informella i kommunikationen genom Facebook och Twitter. Den argumentation som förs är genomgående stängd och Försvarsmakten tycks inte, i den kommunikation som analyserats, ha något intresse av att diskutera sin verksamhet via Facebook eller Twitter. Kommunikationen via de sociala medi--‐ erna uppstår i vissa fall till följd av att en privatperson riktar en fråga till For--‐ svarsmakten. Att påstå att kommunikationen via de sociala medierna därför har en rubbad maktbalans är dock inte rimligt. Den information som sprids och dis--‐ kuteras via Försvarsmaktens Facebooksida sker fortfarande till största del på Försvarsmaktens villkor. Däremot är det rimligt att påstå att Facebooksidan fun--‐ gerar som en ny mötesplats för de som har ett intresse av Försvarsmakten och deras verksamhet. Nyckelord: textuell kommunikation, sociala medier, fallstudie, myndighet, bruks--‐ textanalys.