Dissertations / Theses on the topic 'User-generated content analysis'

This paper aims to show how User Generated Content can create value for Smart Tourism Destinations. Applying the analysis on 5 different cases in the region of Stockholm to derive patterns and opportunities of value creation generated by UGC in tourism. Findings of this paper is also discussed in terms of improving decision making, possibilities of new business models and importance of technological improvements on STD’s. Finally, thoughts on models are presented for researchers and practitioners that might be interested in exploitation of UGC in the context of information-intensive industries and mainly in Tourism.

In recent years emotion-related applications like smartphone apps that document and analyse the emotions of the user, have become very popular. But research also can deal with human emotions in a very technology-driven approach. Thus space-related emotions are of interest as well which can be visualised cartographically and can be captured in different ways. The research project of this dissertation deals with the extraction of georeferenced emotions from the written language in the metadata of Flickr and Panoramio photos, thus from user-generated content, as well as with their modelling and visualisation. Motivation is the integration of an emotional component into location-based services for tourism since only factual information is considered thus far although places have an emotional impact. The metadata of those user-generated photos contain descriptions of the place that is depicted within the respective picture. The words used have affective connotations which are determined with the help of emotional word lists. The emotion that is associated with the particular word in the word list is described on the basis of the two dimensions ‘valence’ and ‘arousal’. Together with the coordinates of the respective photo, the extracted emotion forms a georeferenced emotion. The algorithm that was developed for the extraction of these emotions applies different approaches from the field of computer linguistics and considers grammatical special cases like the amplification or negation of words. The algorithm was applied to a dataset of Flickr and Panoramio photos of Dresden (Germany). The results are an emotional characterisation of space which makes it possible to assess and investigate specific features of georeferenced emotions. These features are especially related to the temporal dependence and the temporal reference of emotions on one hand; on the other hand collectively and individually perceived emotions have to be distinguished. As a consequence, a place does not necessarily have to be connected with merely one emotion but possibly also with several. The analysis was carried out with the help of different cartographic visualisations. The temporal occurrence of georeferenced emotions was examined detailed. Hence the dissertation focuses on fundamental research into the extraction of space-related emotions from georeferenced user-generated content as well as their visualisation. However as an outlook, further research questions and core themes are identified which arose during the investigations. This shows that this subject is far from being exhausted<br>In den letzten Jahren sind emotionsbezogene Anwendungen, wie Apps, die die Emotionen des Nutzers dokumentieren und analysieren, sehr populär geworden. Ebenfalls in der Forschung sind Emotionen in einem sehr technologiegetriebenen Ansatz ein Thema. So auch ortsbezogene Emotionen, die sich somit kartographisch darstellen lassen und auf verschiedene Art und Weisen gewonnen werden können. Das Forschungsvorhaben der Dissertation befasst sich mit der Extraktion von georeferenzierten Emotionen aus geschriebener Sprache unter Verwendung von Metadaten verorteter Flickr- und Panoramio-Fotos, d.h. aus nutzergenerierten Inhalten, sowie deren Modellierung und Visualisierung. Motivation hierfür ist die Einbindung einer emotionalen Komponente in ortsbasierte touristische Dienste, da diese bisher nur faktische Informationen berücksichtigen, obwohl Orte durchaus eine emotionale Wirkung haben. Die Metadaten dieser nutzergenerierten Inhalte stellen Beschreibungen des auf dem Foto festgehaltenen Ortes dar. Die dafür verwendeten Wörter besitzen affektive Konnotationen, welche mit Hilfe emotionaler Wortlisten ermittelt werden. Die Emotion, die mit dem jeweiligen Wort in der Wortliste assoziiert wird, wird anhand der zwei Dimensionen Valenz und Erregung beschrieben. Die extrahierten Emotionen bilden zusammen mit der geographischen Koordinate des jeweiligen Fotos eine georeferenzierte Emotion. Der zur Extraktion dieser Emotionen entwickelte Algorithmus bringt verschiedene Ansätze aus dem Bereich der Computerlinguistik zum Einsatz und berücksichtigt ebenso grammatikalische Sonderfälle, wie Intensivierung oder Negation von Wörtern. Der Algorithmus wurde auf einen Datensatz von Flickr- und Panoramio-Fotos von Dresden angewendet. Die Ergebnisse stellen eine emotionale Raumcharakterisierung dar und ermöglichen es, spezifische Eigenschaften verorteter Emotionen festzustellen und zu untersuchen. Diese Eigenschaften beziehen sich sowohl auf die zeitliche Abhängigkeit und den zeitlichen Bezug von Emotionen, als auch darauf, dass zwischen kollektiv und individuell wahrgenommenen Emotionen unterschieden werden muss. Das bedeutet, dass ein Ort nicht nur mit einer Emotion verbunden sein muss, sondern möglicherweise auch mit mehreren. Die Auswertung erfolgte mithilfe verschiedener kartographischer Visualisierungen. Eingehender wurde das zeitliche Auftreten der ortsbezogenen Emotionen untersucht. Der Fokus der Dissertation liegt somit auf der Grundlagenforschung zur Extraktion verorteter Emotionen aus georeferenzierten nutzergenerierten Inhalten sowie deren Visualisierung. Im Ausblick werden jedoch weitere Fragestellungen und Schwerpunkte genannt, die sich im Laufe der Untersuchungen ergeben haben, womit gezeigt wird, dass dieses Forschungsgebiet bei Weitem noch nicht ausgeschöpft ist

This thesis developed new methods to find useful information from massive customer generated product review data in order to assist customers in decision making. It first examines the distinct features of review text to find useful information about the reviewed product using a number of existing techniques. Then, based upon derived product information, this thesis developed novel methods to assess the review quality in order to find most useful or helpful product reviews for customers.

The present study examines typical features of internet language found in user-generated comments collected from commenting groups from four online magazines aimed at different readerships: (1) adult women (Working Mother and Mothering), (2) adult men (Esquire), (3) young women (Seventeen) and (4) young men (Gameinformer). Approximately 5,000 words from each commenting group were collected, creating a 21,087 word corpus which was analyzed with regard to typographic (emoticons, nonstandard typography of and, personal pronouns you and I) and orthographic features (abbreviations, acronyms) as well as syntactic and stylistic features resembling spoken language (contracted forms, ellipsis of subject and/or verb and commenting tone). The results show that adult men wrote the longest comments, followed by adult women, young men and young women in descending order. Furthermore, as for the typical features regarding typography and orthography, it was found that among the four commenting groups, adult men and adult women used them very sparsely, young men used them occasionally and young women used the features most frequently. The analysis of tone showed that adult men mostly used an aggressive or neutral tone, while adult women, young women and young men mostly used a friendly or neutral tone. Young women used an aggressive tone more often than adult women and young men. Moreover, regarding the syntactic and stylistic features, results revealed that the young men were the most frequent users of ellipsis of subject and/or verb, followed by adult women, young women and adult men. Contracted forms were used extensively in the potential places of contractions, regardless of commenting group. Since young men used the ellipsis of subject and/or verb most frequently of all commenting groups and also used the contracted forms in all potential places of contractions, the conclusion is that the young men used a style that is closer to spoken English than the three other commenting groups.<br>Den här studien undersöker språkdrag som är typiska för språk på internet. Det material som har undersöks har hämtats från användarkommentarer i nättidningar som är riktade till fyra olika läsargrupper: (1) kvinnor (Working Mother, Mothering), (2) män (Esquire), (3) unga kvinnor (Seventeen) och (4) unga män (Gameinformer). Cirka 5 000 ord hämtades från kommentarsfälten för varje tidning, vilket resulterade i en korpus som omfattade 21 087 ord totalt. Korpusen analyserades med hänsyn till typografiska språkdrag (smileys, ickestandardiserad stavning av personliga pronomen I och you samt and) och ortografiska språkdrag (förkortningar, akronymer) samt syntaktiska och stilistiska språkdrag som påminner om talspråk (sammandragningar, ellips av subjekt och/eller predikatsverb, tonläge). Resultaten visade att män skrev de längsta kommentarerna, följda av kvinnor, unga män och unga kvinnor i fallande ordning. Vad gäller typiska typografiska och ortografiska språkdrag visar resultatet att de återfanns mycket sparsamt i kvinnornas och männens data, att de återfanns då och då i de unga männens data och att de unga kvinnorna var de som använde dessa språkdrag mest frekvent. Analys av tonläge i användarkommentarerna visade att män oftast använde en aggressiv eller neutral ton, medan kvinnor, unga kvinnor och unga män oftast använde en vänskaplig eller neutral ton. Unga kvinnor använde en aggressiv ton oftare än kvinnor och unga män. Utöver detta visade resultatet att ellips av subjekt och/eller predikatsverb var mest frekvent i de unga männens användarkommentarer, följt av kvinnornas, de unga kvinnornas och männens. Sammandragna former användes näst intill undantagslöst i hela korpusen. Eftersom pojkarna uppvisade mest frekvent användning av ellips av subjekt och/eller predikatsverb samt använde sammandragna former i full utsträckning, kan slutsatsen dras att de unga männens syntax är mer påverkad av engelskt talspråk än syntaxen hos de tre andra kommenterande grupperna.

Since the World Wide Web (Web) became a medium to serve information, its impact on geographic information has been constantly growing. Today the evolution of the bi-directional Web 2.0 has created the phenomenon of User Generated Spatial Content. In this Thesis the focus is into analysing different aspects of this phenomenon from the perspective of a mapping agency and also developing methodologies for meeting the challenges revealed. In this context two empirical studies are conducted. The first examines the spatial dimension of the popular Web 2.0 photo-sharing websites like Flickr, Panoramio, Picasa Web and Geograph, mainly investigating whether such Web applications can serve as sources of spatial content. The findings show that only Web applications that urge users to interact directly with spatial entities can serve as universal sources of spatial content. The second study looks into data quality issues of the OpenStreetMap, a popular wiki-based Web mapping application. Here the focus is on the positional accuracy and attribution quality of the user generated spatial entities. The research reveals that positional accuracy is fit for a number of purposes. On the other hand, the user contributed attributes suffer from inconsistencies. This is mainly due to the lack of a methodology that could help to the formalisation of the contribution process, and thus enhance the overall quality of the dataset. The Thesis explores a formalisation process through an XML Schema for remedying this problem. Finally, the advantages of using vector data in order to enhance interactivity and thus create more efficient and bi-directional Web 2.0 mapping applications is analysed and a new method for vector data transmission over the Web is presented.

E-commerce distribution channel experiences a dramatic development nowadays. France witnessed a rapid growth rate of online sales in recent years and now ranks as the second largest market in Europe in terms of turnover. On E-commerce websites, customer review system is considered as an efficient tool of E-Word of Mouth, enabling users to write recommendations which will influence potential purchasers. This paper discusses the crucial factors of customer review system. Also, a tool to evaluate review system is elaborated with five criteria such as accessibility, quality, design, interaction and control. Ten case studies of French E-commerce firms are presented according to those criteria in qualitative study. Opinions from French customers are collected through online questionnaires in quantitative study. Researches made from supply (firms) and demand (customers) sides show that accessibility and quality are the main concerns for users and often weaknesses in current review systems. Also, the credibility of reviews is questioned by customers. Focusing on these aspects, the paper aims to give suggestions for designing an ideal customer review system to firms in French e-commerce industry.

Conteúdo Gerado por Usuário (CGU) é a denominação dada ao conteúdo criado de forma espontânea por indivíduos comuns, sem vínculos com meios de comunicação. Esse tipo de conteúdo carrega informações valiosas e pode ser explorado por diversas áreas do conhecimento. Muito do CGU é disponibilizado em forma de textos avaliações de produtos, comentários em fóruns sobre filmes e discussões em redes sociais são exemplos. No entanto, a linguagem utilizada em textos de CGU diverge, de várias maneiras, da norma culta da língua, dificultando seu processamento por técnicas de PLN. A linguagem de CGU é fortemente ligada à língua utilizada no cotidiano, contendo, assim, uma grande quantidade de ruídos. Erros ortográficos, abreviações, gírias, ausência ou mau uso de pontuação e de capitalização são alguns ruídos que dificultam o processamento desses textos. Diversos trabalhos relatam perda considerável de desempenho ao testar ferramentas do estado-daarte de PLN em textos de CGU. A Normalização Textual é o processo de transformar palavras ruidosas em palavras consideradas corretas e pode ser utilizada para melhorar a qualidade de textos de CGU. Este trabalho relata o desenvolvimento de métodos e sistemas que visam a (a) identificar palavras ruidosas em textos de CGU, (b) encontrar palavras candidatas a sua substituição, e (c) ranquear os candidatos para realizar a normalização. Para a identificação de ruídos, foram propostos métodos baseados em léxicos e em aprendizado de máquina, com redes neurais profundas. A identificação automática apresentou resultados comparáveis ao uso de léxicos, comprovando que este processo pode ser feito com baixa dependência de recursos. Para a geração e ranqueamento de candidatos, foram investigadas técnicas baseadas em similaridade lexical e word embeddings. Concluiu-se que o uso de word embeddings é altamente adequado para normalização, tendo atingido os melhores resultados. Todos os métodos propostos foram avaliados com base em um córpus de CGU anotado no decorrer do projeto, contendo textos de diferentes origens: fóruns de discussão, reviews de produtos e publicações no Twitter. Um sistema, Enelvo, combinando todos os métodos foi implementado e comparado a um outro sistema normalizador existente, o UGCNormal. Os resultados obtidos pelo sistema Enelvo foram consideravelmente superiores, com taxa de correção entre 67% e 97% para diferentes tipos de ruído, com menos dependência de recursos e maior flexibilidade na normalização.<br>User Generated Content (UGC) is the name given to content created spontaneously by ordinary individuals, without connections to the media. This type of content carries valuable information and can be exploited by several areas of knowledge. Much of the UGC is provided in the form of texts product reviews, comments on forums about movies, and discussions on social networks are examples. However, the language used in UGC texts differs, in many ways, from the cultured norm of the language, making it difficult for NLP techniques to handle them. UGC language is strongly linked to the language used in daily life, containing a large amount of noise. Spelling mistakes, abbreviations, slang, absence or misuse of punctuation and capitalization are some noises that make it difficult to process these texts. Several works report considerable loss of performance when testing NLP state-of-the-art tools in UGC texts. Textual Normalization is the process of turning noisy words into words considered correct and can be used to improve the quality of UGC texts. This work reports the development of methods and systems that aim to (a) identify noisy words in UGC, (b) find candidate words for substitution, and (c) rank candidates for normalization. For the identification of noisy words, lexical-based methods and machine learning ones using deep neural networks were proposed. The automatic identification presented results comparable to the use of lexicons, proving that this process can be done with low dependence of resources. For the generation and ranking of candidates, techniques based on lexical similarity and word embeddings were investigated. It was concluded that the use of embeddings is highly suitable for normalization, having achieved the best results. All proposed methods were evaluated based on a UGC corpus annotated throughout the project, containing texts from different sources: discussion forums, product reviews and tweets. A system, Enelvo, combining all methods was implemented and compared to another existing normalizing system, UGCNormal. The results obtained by the Enelvo system were considerably higher, with a correction rate between 67 % and 97 % for different types of noise, with less dependence on resources and greater flexibility in normalization.

The shift of media from traditional forms to new digital ones has raised the possibility of new kinds of media services, including mobile television. In today¿s communications market, mobile phones are of increasing importance to users and, since mobile devices are connected most of the time, they have a high degree of location independence. The availability of 3G technology and the mobile devices needed to implement mobile television are now established and available. Mobile television is expected to be an important new service that could penetrate the market place and provide new applications, as well as create a market for new players and new investments, if the appropriate price, content and philosophy for content design are found. This research explores the many potential application areas for mobile TV, with a particular focus on advertising. Various organisations that seek success in this market can utilise the potential for advertising on mobile TV. Ultimately, mobile device users are able to use mobile TV for entertainment and information sourcing. However, a number of challenging issues remain to be addressed. The features that appealed to the consumers were studied in this research. Surveys were conducted to obtain an understanding of consumers¿ opinions and needs regarding the mobile TV experience. Many users clearly do like to interact with video content on mobile devices. Interactive mobile TV advertising can benefit users who will be able to use an essentially ¿free¿ mobile TV service, funded by an advertising model. This research proposes an environment for interactive advertising on mobile TV and discussion of an implementation of the proposed designs.

The shift of media from traditional forms to new digital ones has raised the possibility of new kinds of media services, including mobile television. In today's communications market, mobile phones are of increasing importance to users and, since mobile devices are connected most of the time, they have a high degree of location independence. The availability of 3G technology and the mobile devices needed to implement mobile television are now established and available. Mobile television is expected to be an important new service that could penetrate the market place and provide new applications, as well as create a market for new players and new investments, if the appropriate price, content and philosophy for content design are found. This research explores the many potential application areas for mobile TV, with a particular focus on advertising. Various organisations that seek success in this market can utilise the potential for advertising on mobile TV. Ultimately, mobile device users are able to use mobile TV for entertainment and information sourcing. However, a number of challenging issues remain to be addressed. The features that appealed to the consumers were studied in this research. Surveys were conducted to obtain an understanding of consumers' opinions and needs regarding the mobile TV experience. Many users clearly do like to interact with video content on mobile devices. Interactive mobile TV advertising can benefit users who will be able to use an essentially 'free' mobile TV service, funded by an advertising model. This research proposes an environment for interactive advertising on mobile TV and discussion of an implementation of the proposed designs.

The present study draws on recent sociological literature that argues that the study of cognition and culture can benefit from theories of embodied cognition. The concept of semantic internalization is introduced, which is conceptualized as the ability to perceive and articulate the topics that are of most concern to a community as they are manifested in social discourse. Semantic internalization is partly an application of emotional intelligence in the context of community-level discourse. Semantic internalization is measured through the application of Latent Semantic Analysis. Furthermore, it is investigated whether this ability is related to an individual’s social capital and habitus. The analysis is based on data collected from the online review platform yelp.com.

Context: There are two ways that digital media may influence alcohol use. The first is commercial alcohol marketing. The second is user-created alcohol promotion, defined as content distributed through new media that promotes consumption, but independent of commercial marketing. This thesis explores how both types of content promote alcohol, what association there is between exposure and alcohol-related attitudes and behaviour, and the differences between marketing and user-created promotion. Method: A mixed method design was employed, divided into two studies. The first was a content analysis of the design features, topical references, and messages suggested about alcohol in digital marketing and user-created promotion on Facebook, Twitter, and YouTube. The second was a cross-sectional survey with young adults (n = 405). This measured awareness of, and participation with, digital marketing and user-created promotion, and the association with consumption, higher-risk drinking, brand recall, expectancies, and drinking motives. Results: The content analysis found that digital marketing had personalised designs which contained subtle and positive messages about consumption, whereas user-created promotion had simpler designs, displayed little ethical practice, and contained overt messages about higher-risk drinking. The cross-sectional survey found that young adults were aware of, and participating with, both digital marketing and user-created promotion, with exposure greater for the latter. Exposure to both types of content was positively associated with alcohol use, higher-risk consumption, and drinking intentions. User-created promotion had a stronger association with all outcomes than marketing. The association between exposure and consumption, for both types of content, was mediated through drinking motives and expectancies. Conclusion: Young adults are aware of, and participating with, a range of digital marketing and user-created promotion. That such exposure is associated with alcohol-related attitudes and behaviour highlights the potential of new media to influence alcohol consumption. Further research is required to better understand young people’s experience with digital media and the challenges of addressing online health risk messages.

The role of FoodPorn, or food photographs, may aid in forming an image of a culinary destination. Due to the intangible nature of travel experiences, Foodporn can give consumers a pre-taste of a destination. Focusing on the increasing influence of user-generated content on social media, this research advances our understanding of how gazing upon FoodPorn enhances culinary destination consumer-based brand equity. This study developed an extension of a model that measures culinary precinct destination attractiveness, and the role of FoodPorn in influencing perceptions held by Brisbane consumers of South Bank, Sunnybank and Broadbeach as culinary precincts.

The interest of researchers and practitioners in e-Word-of-Mouth has accelerated with the rise of social media platforms. In the last decade it became more apparent that customers are not merely recipients of product and brand information, but also active participants in shaping product and brand perceptions. However, still no comprehensive understanding of customers’ desire to engage in positive and negative e-Word-of-Mouth has been found. Therefore, this study focuses on Customer Engagement in order to provide knowledge on what makes customers engage in e-Word-of-Mouth. We narrow our study to New Product Introductions due to their crucial role in both business success and failure. In order to study this topic, quantitative methodology using big data analysis of around 20 millions tweets in total, with text analysis of around three million items, obtained from Twitter has been applied. The findings indicate that Brand Benefits and Innovativeness Level are two important aspects that impact Attitudinal Engagement which enables Customer Engagement Behavior.

Mestrado em Ciências Empresariais<br>User-generated content on Web 2.0 touristic websites can be important both for clients and companies of the sector. In the present work it were analyzed 503 online reviews, divided into 2769 sentence segments, from 22 restaurants in the Lisbon region, during the summer period 2012, on Tripadvisor.com. Resorting to an exploratory approach, the objective of this work is to identify the presence of attitude (affect, appreciation and judgement), in accordance with the Appraisal Theory. As well as verify the most mentioned attributes and polarity in each segment. Inter-rater agreement between two other evaluators was also checked, for attitude. The results obtained indicate that the dominant attitude is appreciation with positive polarity. Judgement is usually present when the service and Staff are mentioned, while affect is not often cited in this sample. This indicates that users tend to be more objective and less emotional on their restaurants evaluation. The most mentioned attributes were Quality of Food, Staff and Communication and Price, the majority of these had also positive polarity, which indicates that even in times of economic contention, Quality of Food should be the main focus. The inter-rater agreement was satisfactory. It is also concluded that user-generated content can be useful for managers to better understand the motivations, necessities and expectations of their clients and, in that way, focus their communication campaigns, products and services in order to answer these issues.<br>A informação gerada pelos utilizadores em sítios turísticos Web 2.0 pode ser potencialmente importante tanto para clientes como para as empresas do sector. Neste trabalho foram analisados 503 comentários online, posteriormente divididos em 2769 segmentos de frase, provenientes do site Tripadisor.com referentes a 22 restaurantes da região de Lisboa, durante o período Verão de 2012. O objetivo do trabalho passou por, recorrendo a uma abordagem exploratória, identificar os tipos de atitude presente (afeto, apreciação e julgamento), de acordo com a Teoria da Avaliação. Assim como, verificar os tipos de atributos mais mencionados e a polaridade em cada segmento. Aferiu-se ainda a concordância da avaliação de atitude, recorrendo a dois avaliadores externos. Os resultados obtidos indicam que a atitude dominante é a apreciação com polaridade positiva. Julgamento é mencionado geralmente quando se aborda o serviço e o Staff, enquanto afeto foi pouco citado nesta amostra. Isto indica que os utilizadores tendem a ser mais objetivos e menos emocionais nas suas avaliações a restaurantes. Os atributos mais mencionados são Qualidade da Comida, Staff e Comunicação e Preço, todos com maioritariamente avaliação positiva, o que demonstra que mesmo em tempos de contenção a Qualidade da Comida deve continuar a ser a principal aposta. A concordância obtida foi satisfatória. Conclui-se ainda que a análise detalhada de comentários pode ser útil para que os gestores compreendam melhor as motivações, necessidades e expectativas dos seus clientes e dessa forma orientem as suas campanhas de comunicação e os seus produtos e serviços de forma a dar-lhes resposta.

My research aims to investigate the role of Facebook users in the construction of the current “refugee crisis” representations. Although Social Networks Sites are recognized as spaces for expressing the public opinion and engaging in the public discussion, fewer Italian researches are focused on the contemporary refugee crisis representations that emerge from Facebook users. Since this, the focus of my research is to consider and analyze user comments as forms of articulating voices that show a part, often ignored, of the public opinion able to enrich the media representations. The empirical material is collected from 3 Facebook pages of Italian newspapers and their posts about the narration of the refugee crisis. These pages are one of the main common sources where people read news in their everyday life and they also represent a journalism in transition: in fact, they manifest the continuity between traditional forms of journalismand new ones, but at the same time they also represent a new way of expressing opinions through comments. Through the notion of degradation of the public sphere, the purpose is to critically reflect about the role of comments, trying to delineate a public sphere model that allow to include online participatory forms, which can on one side promote informed citizenry and collective reflection, on the other increase opinion radicalization, polarization and the spread of niche audiences. I have chosen a qualitative mixed approach (rather than a quantitative approach with a massive among of data) in order to deepen better discourses, metaphors, images that shape the representation of the migration crisis produced by users through their comments. I have mainly adopted a discursive approach based on discourses’ analysis, sometime integrated with features of critical discourse analysis frequently used to study racism discourse in media such as the press (Van Dijk). The most evident outcome of this preliminary exploration is the re-activation of rhetorics and discourses about migration phenomena previously stabilized in traditional media environment: this can be seen as a consequence of the hybrid media system where new media and traditional ones are integrated and influence each other. A vast part of public perception is the result of images, political discourses and slogans highly produced by the press and televisions. These results can confirm that the limits of the main patterns of journalistic information about migrations are reproduced in the publicperception showed by users comments. Moreover, the polarization that characterized the media representation of the so called “refugee crisis” is reproduced in the users comments, but the negative side find more space. The interesting point is that this polarization shows a cleargeneralization of the migrant figure based on the denial of individual identity. Migrants have the same collective, ethnics and religious identity. In both sides of the representation, individual past, personal identity, single stories are mainly ignored in users comments. The result is the un-personalized people. More in general, the contemporary migration phenomena is perceived as a huge problem to fix or a danger to avoid. Users Comments stabilized and reinforce the idea of migration as an emergency, and not as a structural fact and a normal process in the modern society.<br>La presente ricerca intende approfondire la relazione tra l’utilizzo dei social media e le rappresentazioni mediali della “crisi dei rifugiati” all’interno la cornice teorica di sfera pubblica online. Il caso di studio analizza la pratica dei commenti online come forma di produzione di contenuti, modalità di partecipazione al dibattito ed espressione dell’opinione pubblica che consegnano alla pratica una valenza partecipativa e potenzialmente democratica. La ricerca, dunque, si è posta di analizzare le principali cornici interpretative prodotte dal pubblico relative al fenomeno migratorio. Il fine è comprendere una specifica pratica d’uso quotidiano della rete in grado di ibridare linguaggi e riarticolare significati, definire, riprodurre e/o modificare percezioni dell’immagine del fenomeno migratorio contemporaneo e, al contempo, cogliere le modalità partecipative del dibattito all’interno della sfera pubblica online. Si intende far dialogare due strumenti tradizionalmente utilizzati nello studio della rappresentazione mediale dei fenomeni migratori: l’analisi del contenuto e l’analisi del discorso. Per tutte le analisi svolte è stato utilizzato il software per l’analisi qualitativa dei dati Atlas.ti come supporto per l’organizzazione, la fase di codifica, l’etichettamento e l’analisi dei dati. I risultati delle analisi dei commenti mostrano 5 macro-frame ovvero un insieme di idee connesse tra loro che richiamano preesistenti schemi interpretativi ai quali il pubblico è stato socializzato. Essi riducono la complessità dei fenomeni, forniscono valutazioni morali rispetto ai fatti e agli attori coinvolti, suggeriscono e giustificano l’applicazione di misure straordinarie. All’interno di questi frame è emersa la nota polarizzazione delle opinioni attorno alla crisi migratoria la quale risulta, tuttavia, fortemente sbilanciata verso i frame nei quali viene riproposto un modello generalizzate “noi-loro” che rientra in un processo di de-individuazione simbolica e omogeneizzazione dei gruppi. Un dato da sottolineare riguarda la presenza di commenti costruiti attorno alla razionalizzazione di discriminazioni e alla politicizzazione della crisi migratoria. I contenuti apertamente violenti, d’incitamento all’odio e alla discriminazione razziale non rappresentano, infatti, il dato maggiormente evidente emerso dalle analisi. Siamo di fronte, piuttosto, a un’opinione pubblica fortemente schierata, composta da utenti non esperti che si esprimono riguardo la politicizzazione del diritto all’asilo legittimando misure politiche di chiusura ed esclusione fino a giustificare episodi di intolleranza e violenza ai danni di rifugiati e richiedenti asilo. Sono emersi, inoltre, alcuni degli aspetti maggiormente critici della partecipazione online basata sulla scarsa qualità del dibattito, sul rafforzamento delle posizioni ideologiche, sullo scontro e il conflitto tra utenti e sulla negazione di spazi di confronto che agevolano la diffusione di microclimi d’opinione. Sebbene da una panoramica generale i risultati suggeriscono che i commenti rappresentano degli spazi per il dibattito pubblico offrendo una nuova occasione partecipativa (Graham 2012), l’analisi del dibattito ha mostrato che gli utenti non !9mettono in gioco quella istruzione formale e quelle risorse materiali necessarie per la partecipazione all’ideale di sfera pubblica delineata da Habermas (1989) che porta a considerare le sezioni dei commenti come spazi online distanti dai principali centri istituzionali del processo decisionale: siamo di fronte a spazi pubblici informali con una funzione limitata alle reazioni pubbliche piuttosto che ai processi di apprendimento collettivo (Privitera 2012).

Bitcoin is a cryptocurrency that has started to become interesting in 2017 as a result of the massive hype the cryptocurrency has created around the world, claiming to be the first fully digital coin based on blockchain technology. There are people reacting and commenting everyday about Bitcoin on social networks and online forums. The known social networks that drive significant opinions related with Bitcoin are Twitter and Reddit. It can be said that Bitcoin may attribute much of its growth to the Reddit social forum. This research is dedicated to the study of the application of Sentiment Analysis to the Reddit users’ comments on daily discussion publications from Bitcoin - Reddit community with the goal to find a correlation between these results and Bitcoin’s known indicators. The study aims to analyze the sentiment of the Bitcoin discussion threads, having two main research questions: 1. Do users’ comments of the daily discussion posts have a correlation with the daily indicators of Bitcoin? 2. Do users’ comments of the daily discussion posts have a correlation when the indicators information of Bitcoin are shifted by a few days? The research has permitted to analyze the correlations between the different indicators.<br>A Bitcoin é uma criptomoeda que começou a tornar-se interessante em 2017 através do entusiasmo que criou pelo mundo, afirmando ser a primeira criptomoeda totalmente digital e baseada na tecnologia de Blockchain. Existem pessoas que reagem e comentam diariamente sobre a Bitcoin em redes sociais e fóruns online. As redes sociais que mais impulsionam opiniões relacionadas com a Bitcoin são o Twitter e o Reddit. Podemos afirmar que a Bitcoin pode atribuir grande parte do seu crescimento através da rede social Reddit. A investigação efetuada no contexto desta tese é dedicada ao estudo da aplicação de uma análise de sentimento aos comentários dos utilizadores do Reddit contidos nas publicações de discussão diária da comunidade Bitcoin, com o objetivo de encontrar uma correlação entre os resultados obtidos e os indicadores conhecidos da Bitcoin. O estudo visa analisar o sentimento associado aos tópicos de discussão da comunidade Bitcoin, tendo duas principais questões de investigação: 1. Os comentários dos utilizadores das publicações de discussão diária têm uma correlação com os indicadores diários da Bitcoin? 2. Os comentários dos utilizadores das publicações de discussão diária têm uma correlação com os indicadores de Bitcoin se forem deslocados por alguns dias? A investigação permitiu analisar as correlações entre os diferentes indicadores.

This study examines information disclosure of Facebook users in America and Kenya. The main aim of this study is to find out the specific type of personal and demographic information that individuals disclose on their profiles. The literature review details the nature of social networking sites highlighting studies that have been carried out in the past on these sites. Self-disclosure is also discussed as described in the social penetration theory. An overview of the practice of public relations profession in Kenya is also given. The research study employs the content analysis methodology, with a total of 500 Facebook profiles being analyzed. Results of this study indicate that users disclose a lot of information on their Facebook profiles such as the use of a self-portrait as the main identifying mark on their profiles, their dates of birth, personal information such as religious and political views and education and work information. There are also major gender differences in information disclosure. There are also differences in information disclosure on Facebook between American users and Kenyan users.<br>Department of Journalism

The abundance and ubiquity of user-generated content has opened horizons when it comes to the organization and analysis of vast and heterogeneous data, especially with the increase of quality of the recording devices witnessed nowadays. Most of the activity experienced in social networks today contains audio excerpts, either by belonging to a certain video file or an actual audio clip, therefore the analysis of the audio features present in such content is of extreme importance in order to better understand it. Such understanding would lead to a better handling of ubiquity data and would ultimately provide a better experience to the end-user. The work discussed in this thesis revolves around using audio features to organize and retrieve meaningful insights from user-generated content crawled from social media websites, more particularly data related to concert clips. From its redundancy and abundance (i.e., for the existence of several recordings of a given event), recordings from musical shows represent a very good use case to derive useful and practical conclusions around the scope of this thesis. Mechanisms that provide a better understanding of such content are presented and already partly implemented, such as audio clustering based on the existence of overlapping audio segments between different audio clips, audio segmentation that synchronizes and relates the different cluster’s clips in time, and techniques to infer audio quality of such clips. All the proposed methods use information retrieved from an audio fingerprinting algorithm, used for the synchronization of the different audio files, with methods for filtering possible false positives of the algorithm being also presented. For the evaluation and validation of the proposed methods, we used one dataset made of several audio recordings regarding different concert clips manually crawled from YouTube.

In recent years emotion-related applications like smartphone apps that document and analyse the emotions of the user, have become very popular. But research also can deal with human emotions in a very technology-driven approach. Thus space-related emotions are of interest as well which can be visualised cartographically and can be captured in different ways. The research project of this dissertation deals with the extraction of georeferenced emotions from the written language in the metadata of Flickr and Panoramio photos, thus from user-generated content, as well as with their modelling and visualisation. Motivation is the integration of an emotional component into location-based services for tourism since only factual information is considered thus far although places have an emotional impact. The metadata of those user-generated photos contain descriptions of the place that is depicted within the respective picture. The words used have affective connotations which are determined with the help of emotional word lists. The emotion that is associated with the particular word in the word list is described on the basis of the two dimensions ‘valence’ and ‘arousal’. Together with the coordinates of the respective photo, the extracted emotion forms a georeferenced emotion. The algorithm that was developed for the extraction of these emotions applies different approaches from the field of computer linguistics and considers grammatical special cases like the amplification or negation of words. The algorithm was applied to a dataset of Flickr and Panoramio photos of Dresden (Germany). The results are an emotional characterisation of space which makes it possible to assess and investigate specific features of georeferenced emotions. These features are especially related to the temporal dependence and the temporal reference of emotions on one hand; on the other hand collectively and individually perceived emotions have to be distinguished. As a consequence, a place does not necessarily have to be connected with merely one emotion but possibly also with several. The analysis was carried out with the help of different cartographic visualisations. The temporal occurrence of georeferenced emotions was examined detailed. Hence the dissertation focuses on fundamental research into the extraction of space-related emotions from georeferenced user-generated content as well as their visualisation. However as an outlook, further research questions and core themes are identified which arose during the investigations. This shows that this subject is far from being exhausted.:Statement of Authorship I Acknowledgements II Abstract III Zusammenfassung V Table of Contents VII List of Figures XI List of Tables XIV List of Abbreviations XV 1 Introduction 1 1.1 Motivation 1 1.2 Research Questions 3 1.3 Thesis Structure 4 1.4 Underlying Publications 4 2 State of the Art 6 2.1 Emotions 6 2.1.1 Definitions and Terms 6 2.1.2 Emotion Theories 7 2.1.2.1 James-Lange Theory 9 2.1.2.2 Two-Factor Theory 9 2.1.3 Structuring Emotions 9 2.1.3.1 Dimensional Approaches 10 2.1.3.2 Basic Emotions 11 2.1.3.3 Empirical Similarity Categories 12 2.1.4 Acquisition of Emotions 14 2.1.4.1 Verbal Procedures 14 2.1.4.2 Non-Verbal Procedures 14 2.1.5 Relation between Emotions and Places 15 2.1.6 Emotions in Language 17 2.1.7 Affect Analysis and Sentiment Analysis 20 2.2 User-Generated Content 22 2.2.1 Definition and Characterisation 22 2.2.2 Advantages and Disadvantages 23 2.2.3 Tagging 24 2.2.4 Inaccuracies 28 2.2.5 Flickr and Panoramio 29 2.2.5.1 Flickr 30 2.2.5.2 Panoramio 31 2.3 Related Work on Georeferenced Emotions 32 2.3.1 Emotional Data Resulting from Biometric Measurements 33 2.3.1.1 Bio Mapping 33 2.3.1.2 EmBaGIS 34 2.3.1.3 Ein emotionales Kiezportrait 35 2.3.2 Emotional Data Resulting from Empirical Surveys 35 2.3.2.1 EmoMap 35 2.3.2.2 WiMo 36 2.3.2.3 ECDESUP 37 2.3.2.4 Map of World Happiness 38 2.3.2.5 Emotional Study of Yeongsan River Basin 39 2.3.3 Emotional Data Resulting from User-Generated Content 40 2.3.3.1 Emography 40 2.3.3.2 Twittermood 40 2.3.3.3 Tweetbeat 42 2.3.3.4 Beautiful picture of an ugly place 42 2.3.4 Visualisation in the Related Work 43 3 Methods 45 3.1 Approach for Extracting Georeferenced Emotions from the Metadata of Flickr and Panoramio Photos 45 3.2 Implemented Algorithm 45 3.3 Grammatical Special Cases 47 3.3.1 Degree Words 48 3.3.2 Negation 52 3.3.2.1 Syntactic Negation in English Language 55 3.3.2.2 Syntactic Negation in German Language 57 3.3.3 Modification of Words Affected by Grammatical Special Cases 60 4 Visualisation and Analysis of Extracted Georeferenced Emotions 62 4.1 Data Basis 62 4.2 Density Maps 67 4.3 Inverse Distance Weight 71 4.4 3D Visualisation 73 4.5 Choropleth Mapping 74 4.6 Point Symbols 78 4.7 Impact of Considering Grammatical Special Cases 80 5 Investigation in Temporal Aspects 85 5.1 Annually Occurrence of Emotions 85 5.2 Periodic Events 87 5.3 Single Events 91 5.4 Dependence of Georeferenced Emotions on Different Periods of Time 93 5.4.1 Seasons 95 5.4.2 Months 96 5.4.3 Weekdays 98 5.4.4 Times of Day 99 5.5 Potentials and Limits of Temporal Analyses 99 6 Discussion 100 6.1 Evaluation 100 6.2 Weaknesses and Problems 102 7 Conclusions and Outlook 105 7.1 Answers to the Research Questions 105 7.2 Outlook and Future Work 107 8 Bibliography 112 Appendices XVI<br>In den letzten Jahren sind emotionsbezogene Anwendungen, wie Apps, die die Emotionen des Nutzers dokumentieren und analysieren, sehr populär geworden. Ebenfalls in der Forschung sind Emotionen in einem sehr technologiegetriebenen Ansatz ein Thema. So auch ortsbezogene Emotionen, die sich somit kartographisch darstellen lassen und auf verschiedene Art und Weisen gewonnen werden können. Das Forschungsvorhaben der Dissertation befasst sich mit der Extraktion von georeferenzierten Emotionen aus geschriebener Sprache unter Verwendung von Metadaten verorteter Flickr- und Panoramio-Fotos, d.h. aus nutzergenerierten Inhalten, sowie deren Modellierung und Visualisierung. Motivation hierfür ist die Einbindung einer emotionalen Komponente in ortsbasierte touristische Dienste, da diese bisher nur faktische Informationen berücksichtigen, obwohl Orte durchaus eine emotionale Wirkung haben. Die Metadaten dieser nutzergenerierten Inhalte stellen Beschreibungen des auf dem Foto festgehaltenen Ortes dar. Die dafür verwendeten Wörter besitzen affektive Konnotationen, welche mit Hilfe emotionaler Wortlisten ermittelt werden. Die Emotion, die mit dem jeweiligen Wort in der Wortliste assoziiert wird, wird anhand der zwei Dimensionen Valenz und Erregung beschrieben. Die extrahierten Emotionen bilden zusammen mit der geographischen Koordinate des jeweiligen Fotos eine georeferenzierte Emotion. Der zur Extraktion dieser Emotionen entwickelte Algorithmus bringt verschiedene Ansätze aus dem Bereich der Computerlinguistik zum Einsatz und berücksichtigt ebenso grammatikalische Sonderfälle, wie Intensivierung oder Negation von Wörtern. Der Algorithmus wurde auf einen Datensatz von Flickr- und Panoramio-Fotos von Dresden angewendet. Die Ergebnisse stellen eine emotionale Raumcharakterisierung dar und ermöglichen es, spezifische Eigenschaften verorteter Emotionen festzustellen und zu untersuchen. Diese Eigenschaften beziehen sich sowohl auf die zeitliche Abhängigkeit und den zeitlichen Bezug von Emotionen, als auch darauf, dass zwischen kollektiv und individuell wahrgenommenen Emotionen unterschieden werden muss. Das bedeutet, dass ein Ort nicht nur mit einer Emotion verbunden sein muss, sondern möglicherweise auch mit mehreren. Die Auswertung erfolgte mithilfe verschiedener kartographischer Visualisierungen. Eingehender wurde das zeitliche Auftreten der ortsbezogenen Emotionen untersucht. Der Fokus der Dissertation liegt somit auf der Grundlagenforschung zur Extraktion verorteter Emotionen aus georeferenzierten nutzergenerierten Inhalten sowie deren Visualisierung. Im Ausblick werden jedoch weitere Fragestellungen und Schwerpunkte genannt, die sich im Laufe der Untersuchungen ergeben haben, womit gezeigt wird, dass dieses Forschungsgebiet bei Weitem noch nicht ausgeschöpft ist.:Statement of Authorship I Acknowledgements II Abstract III Zusammenfassung V Table of Contents VII List of Figures XI List of Tables XIV List of Abbreviations XV 1 Introduction 1 1.1 Motivation 1 1.2 Research Questions 3 1.3 Thesis Structure 4 1.4 Underlying Publications 4 2 State of the Art 6 2.1 Emotions 6 2.1.1 Definitions and Terms 6 2.1.2 Emotion Theories 7 2.1.2.1 James-Lange Theory 9 2.1.2.2 Two-Factor Theory 9 2.1.3 Structuring Emotions 9 2.1.3.1 Dimensional Approaches 10 2.1.3.2 Basic Emotions 11 2.1.3.3 Empirical Similarity Categories 12 2.1.4 Acquisition of Emotions 14 2.1.4.1 Verbal Procedures 14 2.1.4.2 Non-Verbal Procedures 14 2.1.5 Relation between Emotions and Places 15 2.1.6 Emotions in Language 17 2.1.7 Affect Analysis and Sentiment Analysis 20 2.2 User-Generated Content 22 2.2.1 Definition and Characterisation 22 2.2.2 Advantages and Disadvantages 23 2.2.3 Tagging 24 2.2.4 Inaccuracies 28 2.2.5 Flickr and Panoramio 29 2.2.5.1 Flickr 30 2.2.5.2 Panoramio 31 2.3 Related Work on Georeferenced Emotions 32 2.3.1 Emotional Data Resulting from Biometric Measurements 33 2.3.1.1 Bio Mapping 33 2.3.1.2 EmBaGIS 34 2.3.1.3 Ein emotionales Kiezportrait 35 2.3.2 Emotional Data Resulting from Empirical Surveys 35 2.3.2.1 EmoMap 35 2.3.2.2 WiMo 36 2.3.2.3 ECDESUP 37 2.3.2.4 Map of World Happiness 38 2.3.2.5 Emotional Study of Yeongsan River Basin 39 2.3.3 Emotional Data Resulting from User-Generated Content 40 2.3.3.1 Emography 40 2.3.3.2 Twittermood 40 2.3.3.3 Tweetbeat 42 2.3.3.4 Beautiful picture of an ugly place 42 2.3.4 Visualisation in the Related Work 43 3 Methods 45 3.1 Approach for Extracting Georeferenced Emotions from the Metadata of Flickr and Panoramio Photos 45 3.2 Implemented Algorithm 45 3.3 Grammatical Special Cases 47 3.3.1 Degree Words 48 3.3.2 Negation 52 3.3.2.1 Syntactic Negation in English Language 55 3.3.2.2 Syntactic Negation in German Language 57 3.3.3 Modification of Words Affected by Grammatical Special Cases 60 4 Visualisation and Analysis of Extracted Georeferenced Emotions 62 4.1 Data Basis 62 4.2 Density Maps 67 4.3 Inverse Distance Weight 71 4.4 3D Visualisation 73 4.5 Choropleth Mapping 74 4.6 Point Symbols 78 4.7 Impact of Considering Grammatical Special Cases 80 5 Investigation in Temporal Aspects 85 5.1 Annually Occurrence of Emotions 85 5.2 Periodic Events 87 5.3 Single Events 91 5.4 Dependence of Georeferenced Emotions on Different Periods of Time 93 5.4.1 Seasons 95 5.4.2 Months 96 5.4.3 Weekdays 98 5.4.4 Times of Day 99 5.5 Potentials and Limits of Temporal Analyses 99 6 Discussion 100 6.1 Evaluation 100 6.2 Weaknesses and Problems 102 7 Conclusions and Outlook 105 7.1 Answers to the Research Questions 105 7.2 Outlook and Future Work 107 8 Bibliography 112 Appendices XVI

碩士<br>淡江大學<br>國際企業學系碩士班<br>106<br>Product-related information generated from both marketers and consumers flow freely on the internet, it is persistent and accessible through various channels, conjunction with consumers rely highly on opinion leader and other peer consumers’ opinion and recommendation, these phenomena formed the powerful impact of brand-related user-generated content on product consideration and consumers’ purchase decision. According to UGC contains two major elements that affect information adoption, which are information quality and source credibility, therefore firms should be savvy to exploit those free marketing contents generated by their consumer as the firms’ producer-generated content effectively. In Twitter, the message or tweet is conveyed to the audience instantly and its basic features allow other users to interact with that tweet, therefore the participant with a large number of followers possesses the ability of information diffusion to massive recipients and tends to receive a high level of engagement. The conceptual model of UGC transformation to PGC process retrieved from the empirical study is developed, then NodeXL and social network analysis method are applied to investigate the effect of diagnostic metrics on the UGC transformation process. The valuable evidences obtained from this study, not only answered the research questions but also demonstrate the pattern of practical marketing practice that any practitioners or firm can adopt in order to improve their online marketing performance

Increased consumer expectations and diverse consumer needs have led to consumers turning to the internet as a communication medium to share their positive and negative experiences, feelings, and level of satisfaction with others. The primary objective of this study was to explore the nature of user-generated content, in terms of consumer complaint and compliment behaviour related to the seven OEM brands in South Africa. The study focused on the South African automotive industry, and the consumer behaviour associated with post-purchase responses on an online consumer advocacy platform, called Hellopeter.com. A qualitative research design, making use of a content analysis, was employed. In total, 185 user-generated content items were collected from Hellopeter.com for the data-collection period, and 176 complaints and 9 compliments were analysed. The results of this study indicated that the nature of user-generated content related to the seven OEM brands in South Africa could be categorised according to eight broad themes: seven themes were specifically related to consumer complaint behaviour, while one theme was associated with consumer compliment behaviour. The top consumer complaints and compliments related to the OEM brands on Hellopeter.com were also determined. The nature of the user-generated content showed various similarities to that of the general services industry but proved to be unique to the automotive industry. Specific recommendations were made and included that open communication channels and proper feedback systems should be put in place to satisfy consumer needs, and that the findings of the study should be used as a benchmark to monitor and improve consumer complaint and compliment behaviour, as a means to create loyal consumers<br>Business Management<br>M. Com (Business Management)

Yes<br>This study uncovers hotel brand positioning and competitive landscape mapping by text-mining user-generated content (UGC). Rather than relying on a single dimension of consumer evaluation, the current study detects brand attributes by using both customer preferences as well as perceptual performance to develop meaningful insights. For this, the study combines content analysis and repertory grid analysis (RGA) to answer three key research issues. 111,986 hotel reviews from two biggest Chinese cities are used to explore and visualize the competitive landscape of six selected hotel brands across three hotel categories. Findings from the study will not only advance the existing literature on brand positioning and competitive landscape mapping but also help practitioners in developing brand positioning strategies to fight competitors within and across hotel categories.

The issue of ‘race’ and policing has generated considerable public controversy. I draw the work of Norman Fairclough in analyzing online public comments responding to three Winnipeg incidents from the summer of 2008: the detainment of Robert Wilson, the inquest into the death of Matthew Dumas and the tasering death of Michael Langan. My main research questions are 1) what characterises these discourses? 2) what processes of social struggle are evident? and 3) what can this tell us about power relations and ideology in society? The analysis of 3342 comments demonstrates power dynamics in discursive struggles over the definition of the relationship between racialized group-members and the police. Specifically, a conservative discursive formation was found to have three interrelated ‘stages’: support for the police, denial of racism and mediating discourses of responsibilization/criminalization. The conclusion considers how a transformative discourse of racialized policing might mitigate prevailing justifications of racial privilege and inequality.

A thesis submitted in partial fulfillment of the requirements for the degree of Doctor in Information Management, specialization in Geographic Information Systems<br>Traditional geographic information has been produced by mapping agencies and corporations, using high skilled people as well as expensive precision equipment and procedures, in a very costly approach. The production of land use and land cover databases are just one example of such traditional approach. On the other side, The amount of Geographic Information created and shared by citizens through the Web has been increasing exponentially during the last decade, resulting from the emergence and popularization of technologies such as the Web 2.0, cloud computing, GPS, smart phones, among others. Such comprehensive amount of free geographic data might have valuable information to extract and thus opening great possibilities to improve significantly the production of land use and land cover databases. In this thesis we explored the feasibility of using geographic data from different user generated spatial content initiatives in the process of land use and land cover database production. Data from Panoramio, Flickr and OpenStreetMap were explored in terms of their spatial and temporal distribution, and their distribution over the different land use and land cover classes. We then proposed a conceptual model to integrate data from suitable user generated spatial content initiatives based on identified dissimilarities among a comprehensive list of initiatives. Finally we developed a prototype implementing the proposed integration model, which was then validated by using the prototype to solve four identified use cases. We concluded that data from user generated spatial content initiatives has great value but should be integrated to increase their potential. The possibility of integrating data from such initiatives in an integration model was proved. Using the developed prototype, the relevance of the integration model was also demonstrated for different use cases.

<div>Digitization has led to emergence of many platforms-based markets. In this dissertation I focus on three different design problems in these markets. The first essay relates to augmented-reality platforms. Pok\'emon Go, an augmented-reality technology-based game, garnered tremendous public interest upon release with an average of 20 million active daily users. The game combines geo-spatial elements with gamification practices to incentivize user movement in the physical world. This work examines the potential externalities that such incentives may have on associated businesses. Particularly, we study the impact of Pok\'emon Go on local restaurants using online reviews as a proxy for consumer engagement and perception. We treat the release of Pok\'emon Go as a natural experiment and study the post-release impact on the associated restaurants. We find that restaurants located near an in-game artifact do indeed observe a higher level of consumer engagement and a more positive consumer perception as compared with those that have no in-game artifacts nearby. In addition, we find that the heterogeneous characteristics of the restaurants moderate the effect significantly. To the best of our knowledge, this study is the first to examine the economic implications of augmented-reality applications. Thereby, our research lays the foundations for how augmented-reality games affect consumer economic behavior. This work also builds insights into the potential value of such associations for business owners and policymakers. </div><div><br></div><div>The second essay focuses on the platform design problem in sponsored seaerch ad-market.Recent advances in technology have reduced frictions in various markets. In this research, we specifically investigate the role of frictions in determining the efficiency and bidding behavior in a generalized second price auction (GSP) -- the most preferred mechanism for sponsored search advertisements. First, we simulate computational agents in the GSP setting and obtain predictions for the metrics of interest. Second, we test these predictions by conducting a human-subject experiment. We find that, contrary to the theoretical prediction, the lower-valued advertisers (who do not win the auction) substantially overbid. Moreover, we find that the presence of market frictions moderates this phenomenon and results in higher allocative efficiency. These results have implications for policymakers and auction platform managers in designing incentives for more efficient auctions.</div><div><br></div><div>The third essay is about user-generated content platforms. These platform utilize various gamification strategies to incentivize user contributions. One of the most popular strategy is to provide platform sponsorships like a special status. Previous literature has extensively studied the impact of having these sponsorships user contributions. We specifically focus on the impact of losing such elite status. Once their contributions to the platform reduce in volume, elite users lose status. Using a unique empirical strategy we show that users continue to contribute high quality reviews, even though they lose their status. We utilize NLP to extract various review characteristics including sentiment and topics. Using an empirical strategy, we find that losing status does not modify the topic of the reviews written by the users, on average. </div><div><br></div>

This dissertation aims to comprehend the impact of deploying user-generated content (UGC) campaigns on consumers’ perceptions of brand coolness and consumer brand engagement. The trending concept of coolness in the beauty industry is studied through electronic word of mouth to understand if brands encouraging their users to post about their brand experiences leads to consumers perceiving them as cool and engaging more positively through those publications. The methodology in use is a netnography, along with a sentiment analysis technique. The analysis consisted in observing the interactions, incited by a user-generated content campaign led by a prestigious beauty brand - Drunk Elephant, between the brand and its online brand community on the social network Instagram for one year to avoid seasonal phenomena. The comments were retrieved using a text-mining tool and analyzed through Natural Language Processing according to their sentiment polarity, and trending topics identified. The data retrieved from the year of 2019 amounted to 67 321 interactions. Results show consumers’ perceptions of coolness can be positively influenced by adopting UGC campaigns, which can also lead to positive consumer brand engagement. Not only do these campaigns generate brand awareness, but they stimulate brand community’s expansion and influence consumers’ perceptions towards the brand. Beauty brands seeking to grow their status of coolness and consumer interactions should consider implementing user-generated content campaigns, as keeping up with the trends in the market is not only regarded as cool but is necessary to remain relevant in the ever-changing marketplace beauty has proven itself to be.<br>Esta dissertação visa entender o impacto da utilização de campanhas de conteúdo gerado pelos utilizadores nas perceções dos consumidores da coolness de uma marca e interações entre marca e consumidores. A tendência coolness na indústria da beleza é examinada através de electronic word-of-mouth para compreender se encorajar os utilizadores a partilhar conteúdo sobre as suas experiências com as marcas, os leva a pensar na marca como cool e a interagir mais com essas publicações. A metodologia usada é uma análise netnográfica em conjunto com uma técnica de análise sentimental. A análise foi conduzida sob interações textuais, incitadas pela campanha da marca de prestígio de beleza – Drunk Elephant, entre a marca e a sua comunidade online na rede social Instagram durante um ano para evitar fenómenos sazonais. Os comentários foram extraídos por text mining e analisados através de processamento de linguagem natural, tendo em conta a polaridade do seu sentimento, e tópicos mais frequentes identificados. Os dados retirados do ano de 2019 totalizaram 67 321 interações. Os resultados demonstram que as perceções de coolness do consumidor podem ser positivamente influenciadas adotando o uso destas campanhas e podem conduzir a interações positivas. Não só estas campanhas criam visibilidade para a marca, como encorajam a expansão da comunidade da marca e influenciam as perceções dos seus consumidores. Marcas na indústria da beleza que procuram aumentar a sua coolness e interações com os consumidores devem considerar implementar campanhas de conteúdo gerado pelos utilizadores, de maneira a manter-se atuais num mercado em constante transformação.

In the recent past, hardly anyone could predict this course of GIS development. GIS is moving from desktop to cloud. Web 2.0 enabled people to input data into web. These data are becoming increasingly geolocated. Big amounts of data formed something that is called "Big Data". Scientists still don't know how to deal with it completely. Different Data Mining tools are used for trying to extract some useful information from this Big Data. In our study, we also deal with one part of these data - User Generated Geographic Content (UGGC). The Panoramio initiative allows people to upload photos and describe them with tags. These photos are geolocated, which means that they have exact location on the Earth's surface according to a certain spatial reference system. By using Data Mining tools, we are trying to answer if it is possible to extract land use information from Panoramio photo tags. Also, we tried to answer to what extent this information could be accurate. At the end, we compared different Data Mining methods in order to distinguish which one has the most suited performances for this kind of data, which is text. Our answers are quite encouraging. With more than 70% of accuracy, we proved that extracting land use information is possible to some extent. Also, we found Memory Based Reasoning (MBR) method the most suitable method for this kind of data in all cases.

This diploma thesis discusses the image of menstruation in the mass media and in user generated content. Menstruation is a biological process that is essential for reproduction. It not only plays a specific role from the biological point of view, but also from the cultural aspect. Menstruation has remained a taboo and been stigmatised for many years. Today this topic is no longer completely taboo, but some stigma and stereotypes remain. This diploma thesis is based on the theory of social construction of reality, where the mass media is considered one element involved in the construction of this reality. The aim of this work is to analyse the image of menstruation in the mass media created by professionals and compare it with the image of this topic in user generated content created by the public on social networks. Two research methods are used to obtain a comprehensive view of the specific issue - a quantitative content analysis and qualitative framework analysis. The quantitative content analysis is used to establish the frequency and occurrence of the topic of menstruation within mass media. The framework analysis is used to establish how this topic is represented and framed in the mass media and in user generated content.

The thesis is focused on the subject of genres in the context of new media, specializing on the biggest Internet online video sharing platform, YouTube. Considering the interactivity of YouTube as a kind of social media site and making contact between the creator and his audience, the genre composition of YouTube channels directly reflects the preferences of the specific audience. The main objective of this thesis is to make an analysis of genres on a sample of five selected professional Czech YouTube channels. The sample is based on current order of channels according to the number of subscribers (namely ViralBrothers, MenT, GEJMR, Jirka Král and Hoggy). After the introduction of how professional YouTube channels operate follows the theoretical part of the thesis defining the basic ideas, which are important to set the methodological framework of the analysis. The analytical part consists of presenting the results of the genre analysis, categorizing genres into a system and defining genres based on their characteristics.

In the Industry 4.0, modern organizations are characterized by an extensive digitalization and use of Information Technology (IT). Even though there are significant advantages in such a technological progress, a noteworthy drawback is represented by cyber risks, whose occurrence dramatically increased over the last years. The information technology literature has shown great interested toward the topic, identifying mainly technical solutions to face these emerging risks. Nonetheless, cyber risks cause business disruption and damages to tangible and intangible corporate assets and require a major integration between technical solutions and a strategic management. Recently, the risk management domain and the supply chain literature have provided studies about how an effective cyber risk management process should be planned, to improve organizational resilience and to prevent financial drawbacks. However, the aforementioned studies are mainly theoretical and there is still a significant lack of empirical studies in the management literature, measuring the potential effects of cyber threats within single companies, and along networks of relationships, in a wider supply chain perspective. The present thesis aims at filling some of these gaps through three empirical essays. The first study has implemented a Grounded Theory approach to develop an interview targeting 15 European organizations. Afterwards, the fuzzy set Qualitative Comparative Analysis (fsQCA) has been performed, in order to ascertain how managers perceive cyber risks. Results contradict studies that focus merely on technical solution, and confirm the dynamic capability literature, which highlights the relevance of a major integration among relational, organizational, and technical capabilities when dealing with technological issues. Moreover, the study proposes a managerial framework that draws on the dynamic capabilities view, in order to consider the complexity and dynamism of IT and cyber risks. The framework proposes to implement both technical (e.g. software, insurance, investments in IT assets) and organizational (e.g. team work, human IT resources) capabilities to protect the capability of the company to create value. The second essay extends the investigation of the drawbacks of cyber risks to supply chains. The study conducts a Grounded Theory empirical investigation toward several European organizations that rely on security and risk management standards in order to choose the drivers of systematic IT and cyber risk management (risk assessment, risk prevention, risk mitigation, risk compliance, and risk governance). The evidence gleaned from the interviews have highlighted that investments in supply chain mitigation strategies are scant, resulting in supply chains that perform like they had much higher risk appetite than managers declared. Moreover, it has emerged a general lack of awareness regarding the effects that IT and cyber risks may have on supply operations and relationships. Thus, a framework drawing on the supply chain risk management is proposed, offering a holistic risk management process, in which strategies, processes, technologies, and human resources should be aligned in coherence with the governance of each organization and of the supply chain as a whole. The final result should be a supply chain where the actors share more information throughout the whole process, which guarantees strategic benefits, reputation protection, and business continuity. The third essay draws on the Situational Crisis Communication Theory (SCCT) to ascertain whether and how different types of cyber breaches differently affect the corporate reputation, defined as a multidimensional construct in which perceptions of customers, suppliers, (potential) employees, investors and local communities converge. Data breaches have been categorized into three groups by the literature, meaning intentional and internal to the organization (e.g., malicious employees stealing customers’ data), unintentional and internal to the organization (e.g., incorrect security settings that expose private information), and intentional and external to the organization (e.g., ransomware infecting companies’ software). However, this is among the first study to analyse the different reputational drawbacks these types may cause. Moreover, the study considers that, in the industry 4.0 era, social media analysis may be of paramount importance for organizations to understand the market. In fact, user-generated content (UGC), meaning the content created by users, might help in understanding which dimensions of the corporate have been more attacked after a data breach. In this context, the study implements the Latent Dirichlet Allocation (LDA) automated method, a base model in the family of “topic models”, to extract the reputational dimensions expressed in UGC of a sample of 35 organizations in nine industries that had a data breach incident between 2013 and 2016. The results reveal that in general, after a data breach, three dimensions—perceived quality, customer orientation and corporate performance— are a subject of debate for users. However, if the data breach was intentional ad malicious, users focused more on the role of firms’ human resources management, whereas if users did not identify a responsible, users focused more on privacy drawbacks. The study complements crisis communication research by categorizing, in a data breach context, stakeholders’ perceptions of a crisis. In addition, the research is informative for risk management literature and reputation research, analysing corporate reputation dimensions in a data breach crisis setting.

The diploma thesis describes the application of theoretical concepts of journalism into the environemnt of two most popular music websites Musicserver and iREPORT. In the theoretical basis for the topic are featured the basic characteristics of online journalism and the problematic nature of the term "ideal critic". Furthermore, there are at first theoretically established and afterwards analyzed the concepts of agenda-setting, gatekeeping, objectivity and bias, participative journalism, user generatec content, professionalism, hypertextuality and intertextuality, the usage of social networks, tabloidization and the analysis will also touch upon the issue of name transformation to fit the Czech language. There is also a semiotic and visual analysis of chosen samples performed. On the basis of concepts, the paper intends to explain the functioning of the websites, the aspects that are influencing their content, the factors that have an effect on the selection of the content, the means of working with the audience, the problems in terms of objectivity in the art criticism, the requirements laid upon the editorial staff members and the level of influence upon the readers. Keywords online journalism, theoretical concepts, Musicserver, iREPORT, music, agenda-setting, gatekeeping, objectivity,...