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.

Social endorsements are user-generated endorsements of products or services, such as “likes” and personal collections, in an online social platform. We examine the effect of prior social endorsements on subsequent users’ tendency to endorse or examine a product in a social shopping context, where a social platform connect consumers and enable a collaborative shopping experience. This research consists of two parts. In part I, we identify two ways prior social endorsements can affect subsequent user behavior: as a crowd endorsement, which is an aggregate number of endorsements a product receives for anyone who comes across the product, and as a friend endorsement, which is an endorsement with the endorser’s identity delivered only to the endorser’s friends or followers. Using a panel data of 1656 products on a leading social shopping platform, we quantify the relationship between crowd and friend endorsements and subsequent examination (“click”) and endorsement (“like”) of the products, noting that examination is a private behavior while endorsement is a public behavior. Our results are consistent with the identity signaling theory where identity-conscious consumers converge with the aspiration group (the followers) in their public behavior (e.g. endorsement) and diverge from the avoidance groups (the crowd). We also find differences between public and private behaviors. Moreover, the symbolic nature of social shopping platform trumps the traditional dichotomy of symbolic/functional product attributes. Part II of this study seeks to clarify the underlying mechanism through lab experiments. We hypothesize that consumers’ evaluative attitude, specifically the value-expressive type, moderates the relationship between crowd and friend endorsements and a focal user’s product choice. Our initial results of the second study show support for this idea in the cases when the product choice is not obvious.

Thesis (M.S.)--Kent State University, 2007. / Title from PDF t.p. (viewed July 2, 2007). Advisor: David B. Robins. Keywords: information retrieval, search engine, social bookmarking, tagging, folksonomy, Internet, World Wide Web. Includes survey instrument. Includes bibliographical references (p. 137-141).

Εκατομμύρια ανθρώπων επιλέγουν καθημερινά να χρησιμοποιήσουν τον Παγκόσμιο Ιστό για ένα ευρύ σύνολο δραστηριοτήτων. Ο αριθμός των χρηστών του διαδικτύου αυξάνεται συνεχώς, όπως επίσης και το σύνολο των διαφορετικών δραστηριοτήτων που μπορούν να εκτελεστούν μέσω ιστοσελίδων και υπηρεσιών του διαδικτύου. Οι χρήστες του Παγκόσμιου Ιστού καθημερινά συμμετέχουν σε διάφορες ψηφιακές δραστηριότητες, οριοθετώντας με αυτόν τον τρόπο τη ψηφιακή τους «ζωή». Κάθε χρήστης μπορεί να στείλει μήνυμα με το ηλεκτρονικό ταχυδρομείο, να επικοινωνήσει και να δημιουργήσει σχέσεις με άλλους χρήστες του Παγκόσμιου Ιστού, να επισκεφτεί ιστότοπους για να ικανοποιήσει τις ενημερωτικές του ανάγκες ή να διατηρεί ένα προσωπικό προφίλ σε έναν ή περισσότερους ιστότοπους κοινωνικής δικτύωσης. Παράλληλα, όλο και περισσότεροι χρήστες του Παγκόσμιου Ιστού χρησιμοποιούν τα ηλεκτρονικά καταστήματα για τις αγορές τους, επιτυγχάνοντας την απευθείας σύνδεση της έρευνας αγοράς με την αγορά προϊόντων ή υπηρεσιών, ενώ ο σχολιασμός και οι απόψεις άλλων χρηστών για προϊόντα και υπηρεσίες αποτελεί άλλο ένα αναμφισβήτητο πλεονέκτημα των ηλεκτρονικών καταστημάτων. Αποτέλεσμα της αυξανόμενης δραστηριοποίησης των χρηστών είναι η συνεχής αύξηση του όγκου των κειμενικών δεδομένων που έχουν παραχθεί από χρήστες (user generated text content - UGTC) στις ιστοσελίδες του Παγκόσμιου Ιστού. Οι δικτυακές κοινότητες αυξάνονται συνεχώς σε μέγεθος και αριθμό, ενώ ταυτόχρονα οι ιστότοποι και οι υπηρεσίες του Παγκόσμιου Ιστού προσφέρουν όλο και περισσότερες δυνατότητες στους χρήστες τους για να δημιουργήσουν, να συντηρήσουν και να δημοσιοποιήσουν περιεχόμενο κειμένου που έχει παραχθεί από τους ίδιους. Αποτέλεσμα της αλληλεπίδρασης των χρηστών αλλά και χρηστών και ιστοσελίδων, είναι ότι ένα αρκετά μεγάλο μέρος της διαδικτυακής πληροφορίας με το οποίο έρχεται σε επαφή ο μέσος χρήστης καθημερινά, έχει παραχθεί από άλλους χρήστες και όχι από τους δημιουργούς του ιστοτόπου. Η μελέτη των χαρακτηριστικών του περιεχομένου που έχει παραχθεί από χρήστες είναι κομβικό σημείο σε μια σειρά ερευνητικών πεδίων. Χαρακτηριστικό παράδειγμα αποτελούν οι μελέτες στα πλαίσια του πεδίου της εξόρυξης άποψης (opinion mining), οι οποίες βασίζονται στο περιεχόμενο των χρηστών για να αλιεύσουν τις απόψεις για ένα θέμα ή ένα προϊόν. Μελέτες, όπως οι παραπάνω, είναι ιδιαίτερα χρήσιμες στην ανάπτυξη σύγχρονων εμπορικών εφαρμογών, που θα προσφέρουν στον καταναλωτή δυνατότητα πληρέστερης ενημέρωσης για τις συναλλαγές που πρόκειται να πραγματοποιήσει. Άλλες περιπτώσεις αφορούν στην ανάλυση των γλωσσολογικών χαρακτηριστικών των κειμενικών δεδομένων που έχουν συνταχθεί από χρήστες. Επίσης, η μελέτη των χαρακτηριστικών του περιεχομένου που έχει παραχθεί από χρήστες του Παγκόσμιου Ιστού είναι ιδιαίτερα σημαντική στη μελέτη του social web, καθώς είναι δυνατόν να προκύψουν χρήσιμα συμπεράσματα τόσο για την εξέλιξή του στο χώρο και στο χρόνο, όσο και για την περαιτέρω εξέλιξη του, προσφέροντας στους χρήστες νέες δυνατότητες μέσα από σύγχρονες εφαρμογές που θα αναπτυχθούν. Σε κάθε περίπτωση, το περιεχόμενο των ιστοσελίδων του Παγκόσμιου Ιστού μπορεί θεωρητικά να κατηγοριοποιηθεί σε δυο κατηγορίες: στα δεδομένα των δημιουργών των ιστοσελίδων και στα δεδομένα που προέκυψαν από τους χρήστες των ιστοσελίδων κατά την αλληλεπίδρασή τους με αυτές. Στόχος της παρούσας μελέτης είναι να μελετήσει το κατά πόσο είναι εφικτή και με ποιόν τρόπο η αυτόματη αναγνώριση ύπαρξης ή μη περιεχομένου κειμένου του Παγκόσμιου Ιστού που έχει παραχθεί από χρήστες. Στα πλαίσια της παρούσας μεταπτυχιακής εργασίας θα εντοπιστούν χαρακτηριστικά, τα οποία θα επιτρέψουν τον αυτόματο εντοπισμό των κειμενικών δεδομένων χρηστών σε μια ιστοσελίδα. Γενικά σε μια ιστοσελίδα υπάρχουν τρεις πηγές πληροφοριών, οι οποίες μπορούν να χρησιμοποιηθούν κατά τη διαδικασία προσδιορσμού του user generated content (UGC): το περιεχόμενο της ιστοσελίδας, το περιβάλλον εμφάνισής της (συνδεσιμότητα με άλλες σελίδες και anchor text) και η δομή της, η οποία περιγράφεται από τα html tags (πχ ο τίτλος της σελίδας, οι λέξεις που παρουσιάζονται με bold κλπ). Η προτεινόμενη μεθοδολογία συνίσταται στην εφαρμογή τεχνικών ανάλυσης της ιστοσελίδας που σκοπό έχουν τον καθορισμό ενός συνόλου χαρακτηριστικών γνωρισμάτων της (features). Το σύνολο των γνωρισμάτων αυτών αποτελείται από τρία επιμέρους είδη χαρακτηριστικών γνωρισμάτων, τα γλωσσολογικά χαρακτηριστικά γνωρίσματα (textual features), τα χαρακτηριστικά γνωρίσματα δομής της ιστοσελίδας (Html tags), και τα χαρακτηριστικά γνωρίσματα απεικόνισης ή εμφάνισης της ιστοσελίδας (Visual and Visually Central Features). Από τα καθορισμένα χαρακτηριστικά θα επιλεγούν πειραματικά εκείνα, τα οποία θα συμμετάσχουν αποδοτικότερα στον αλγόριθμο προσδιορισμού για την ύπαρξη user generated text content σε μια ιστοσελίδα. Η αξιολόγηση των αποτελεσμάτων της προτεινόμενης μεθόδου θα πραγματοποιηθεί βάσει ενός συνόλου δεδομένων ελέγχου. Τα δεδομένα ελέγχου αποτελούνται από μια συλλογή ιστοσελίδων, για τις οποίες έχει γίνει έλεγχος για το αν περιέχουν user generated text content. Η διαδικασία αξιολόγησης συνίσταται στην σύγκριση των αποτελεσμάτων του αλγορίθμου που προτείνουμε με τα αποτελέσματα που έχουν παραχθεί από την επεξεργασία του συνόλου δεδομένων ελέγχου. Τα συμπεράσματα που θα προκύψουν μπορούν να χρησιμοποιηθούν για την περεταίρω βελτίωση του αλγορίθμου προσδιορισμού ύπαρξης user generated text content, καθώς και για την αξιοποίηση τους σε τεχνικές ανάλυσης και επεξεργασίας του user generated text content από ιστοσελίδες του Παγκόσμιου Ιστού. / Millions of people every day use the Web for a wide range of activities. The number of Internet users is continuously growing, as well as all the different activities that can be performed through websites and Internet services. Web users daily participate in various digital activities, delimiting in this way their digital "life." Each user can send an e-mail, communicate and establish relationships with other web users, visit websites in order to satisfy his information needs, or keep a personal profile in one or more social networking sites. At the same time, more and more web users use online shopping for their purchases, achieving direct connection of the market research by buying products or services, while the commentations and the views of other users for goods and services is another undoubted advantage of online shops. The users’ increasing activity has as result the continuous raising of the volume data, generated by users (user generated content - UGTC) in Web pages. On-line communities are growing in size and number, while simultaneously websites and web services offer users more and more options, in order to create, preserve and publish text produced by them. Result of the interaction between users and the website users and websites, is that a large part of the online information, in which the user come daily in contact, is produced by others and not by the creators of the website. The study of the characteristics of the content obtained by users is a key point in a series of searching fields. Typical examples are the studies within the field of mining opinion (opinion mining), which are based on the content of users to catch their opinions on a topic or product. Studies such as the above, are particularly useful in the development of modern commercial applications that can offer the consumer better information for his transactions. Other cases concern the analysis of the linguistic characteristics of textual data compiled by users. Also, the study of the content characteristics generated by users of the World Wide Web is particularly important in the study of the social web, as well as it can yield useful results for both the evolution in space and time, and further development providing users with new capabilities through new applications, which are developed. In any case, the Web content could theoretically be categorized in two categories: data of Web pages creators and the data generated by web users when interacting with them. The aim of this study is to examine whether it is feasible and with which way the automatic recognition of the text content on the Web produced by users. In this thesis, it will be identified characteristics that allow the automatic detection of textual data of users to a website and its separation from the content that has been produced by the creators of the website. During the planning and design of the proposed method it will initially be studied the inherent characteristics of different types of websites, which are indicative of the presence of these text content users. It will also be studied the usefulness of linguistic and visual features for recognition textual data users at the site, in order to separate it from the official content, that is from content creators. Generally in a website there are three sources of information that can be used during the process of identifying user generated content (UGC): website content, setting of development (connectivity with other pages and anchor text) and its structure, which is described by the html tags (eg page title, words presented in bold, etc). The proposed methodology is recommended in applying technical analysis website aimed at defining a part of attributes (features). All these attributes consist of three kinds of features, textual features, features of the site structure (Html tags), and imaging features or appearance of the website (Visual and Visually Central Features). From the defined features it will experimentally be selected those, which will efficiently participate in the identification algorithm for the existence user generated text content on a website. The evaluation results of the proposed method will be held considering specific audit data. The audit data consist of a collection of web pages, which have already been checked whether they contain user generated text content. The process evaluation reccommends comparing the results of the proposed algorithm with the results obtained from processing all audit data. The conclusions can be used to further improvement of the identification algorithm existence of user generated text content, as well as to exploit them in technical analysis and processing of user generated text content from Web pages.

This thesis proposes a critical study of the techniques and motives behind modern commodity feminist advertising, focusing on the appropriation of the “young girl” as a symbol of the feminist cause. This evolving trend in advertising, building upon new movements of empowerment and the recent proliferation of the online feminist space, is shifting the logics of consumption by marketing feminist ideology and activism through consumer purchasing power. By prompting consumers to believe that their purchases can make a significant change, companies are developing brand loyalty in their key marketing demographics by using the image and rhetoric of the “young girl” to tap into a term I call “anti-nostalgia,” a nostalgia whereby women leverage the inherent sentimentality of childhood with a constructive understanding and rejection of the destructively sexist climate they experienced to combat these sociocultural conditions for future generations. Joining theoretical research on branding, user-generated content, and the neoliberal ideology of the consumer-citizen, I argue that these advertising campaigns, coupled with online spaces for public interaction and participation, effectively create channels for their target consumers to contribute to this commodified form of activism. In reality, however, these “feminist” purchases are simply forms of consumer self-therapy in a modern political climate of systemic gender discrimination.

Background: The modern day era of the Internet gave birth to the growing phenomenon of Internet memes (IM), a type of online user generated content (UGC) (Gangadharbatla, 2008). Now marketers have begun researching the relationship between UGC and consumer based brand equity (CBBE) (Christodoulides et.al, 2012; Rachna and Khajuria, 2017). The problem discussion presents the issue of the diminishing control of brand equity due to the rise of UGC and lack of research on how to manage its influence (Morrison et al., 2013). Leading to the purpose of this thesis, which is to determine the impact Internet memes have on consumer based brand equity. Literature: The review presents two leading contributors to the field, Aaker’s (1991) framework on the different dimensions of CBBE and Keller’s (1993) definition of the concept. In addition recent studies on UGC and brand equity provided the basis for hypothesis development. Method: This thesis assumed a deductive research approach, developing the hypothesis from current literature in the field. A quantitative study, that utilized an explanatory research approach, because it best suited the experimental design. As for the data collection method, surveys were considered (Saunders et al., 2016), which the authors designed as a self-completion questionnaire and pre-tested (Bryman and Bell, 2015). Convenience sampling was chosen to select participants (Hernon, 2004). Construct and content validity was used along with homogeneity and stability to control reliability and measure the quality of research instruments (Bryman and Bell, 2015). SPSS version 25 was used to conduct all statistical analyses. Results: Four hypotheses were developed, to measure the effect of the independent variable IM on each of the four CBBE dimensions. To summarize, three of the hypothesis (H1, H3 and H4) were rejected as the difference in the means are not significant enough and can be explained by chance. The effect on brand association (H2) was however found to be significant therefore H2 was accepted.

O sucesso da mineração de opiniões para processar automaticamente grandes quantidades de conteúdo opinativo disponíveis na Internet tem sido demonstrado como uma solução de baixa latência e mais barata para a análise de opinião pública. No presente trabalho foi investigado se é possível prever variações de intenção de voto com base em séries temporais de sentimento extraídas de comentários de notícias, utilizando três eleições brasileiras como estudo de caso. As contribuições deste estudo de caso são: a) a comparação de duas abordagens para a mineração de opiniões em conteúdo gerado por usuários em português do Brasil; b) a proposta de dois tipos de atributos discriminantes para representar o sentimento em relação a candidatos políticos a serem usados para a previsão, c) uma abordagem para prever variações de intenção de voto que é adequada para cenários de dados esparsos. Foram desenvolvidos experimentos para avaliar a influência dos atributos discriminantes propostos em relação a acurácia da previsão, e suas respectivas preparações. Os resultados mostraram uma acurácia de 70% na previsão de variações de intenção de voto positivas e negativas. Estas contribuições são importantes passos em direção a um framework que é capaz de combinar opiniões de diversas fontes para encontrar a representatividade de uma população alvo, de modo que se possa obter previsões mais confiáveis. / The success of opinion mining for automatically processing vast amounts of opinionated content available on the Internet has been demonstrated as a less expensive and lower latency solution for gathering public opinion. In this work, we investigate whether it is possible to predict variations in vote intention based on sentiment time series extracted from news comments, using three Brazilian elections as case study. The contributions of this case study are: a) the comparison of two approaches for opinion mining in user-generated content in Brazilian Portuguese; b) the proposition of two types of features to represent sentiment behavior towards political candidates that can be used for prediction, c) an approach to predict polls vote intention variations that is adequate for scenarios of sparse data. We developed experiments to assess the influence on the forecasting accuracy of the proposed features, and their respective preparation. Our results display an accuracy of 70% in predicting positive and negative variations. These are important contributions towards a more general framework that is able to blend opinions from several different sources to find representativeness of the target population, and make more reliable predictions.

With emerging content forums blurring the distinctions between journalistic paradigms, this study helps illuminate those which best promote alternative practice. A content analysis of Deepwater Horizon oil spill coverage compared three platforms for online citizen journalism: corporate (CNN iReport), alternative (Indymedia), and independent blogs. News stories were coded for sources, links, author-reader interaction, mobilizing information, tone for the liable parties' ability and intent in handling the disaster, and contestation of official information. Results show that Indymedia was the most alternative in inclusion of mobilizing information, critical tone, contestation of mainstream versions, ratio of alternative links to mainstream, and total usage of alternative sources. iReport engendered the greatest rates of community via interaction, while also averaging the highest ratio of alternative sources. The blogs split on nearly all metrics, as one rated highly in every category and the other near last. This analysis determines which platforms are most likely to cultivate disaster news that stands as alternative to, and not extension of, the mainstream. This study makes a contribution to the theory of alternative media and is the first to compare citizen journalism sites against one another in measuring their adherence to the alternative paradigm, and its examination of CNN's citizen-report model also represents a novel contribution. The findings discussed may help direct citizens as they reach out to online communities in times of disaster.

O sucesso da mineração de opiniões para processar automaticamente grandes quantidades de conteúdo opinativo disponíveis na Internet tem sido demonstrado como uma solução de baixa latência e mais barata para a análise de opinião pública. No presente trabalho foi investigado se é possível prever variações de intenção de voto com base em séries temporais de sentimento extraídas de comentários de notícias, utilizando três eleições brasileiras como estudo de caso. As contribuições deste estudo de caso são: a) a comparação de duas abordagens para a mineração de opiniões em conteúdo gerado por usuários em português do Brasil; b) a proposta de dois tipos de atributos discriminantes para representar o sentimento em relação a candidatos políticos a serem usados para a previsão, c) uma abordagem para prever variações de intenção de voto que é adequada para cenários de dados esparsos. Foram desenvolvidos experimentos para avaliar a influência dos atributos discriminantes propostos em relação a acurácia da previsão, e suas respectivas preparações. Os resultados mostraram uma acurácia de 70% na previsão de variações de intenção de voto positivas e negativas. Estas contribuições são importantes passos em direção a um framework que é capaz de combinar opiniões de diversas fontes para encontrar a representatividade de uma população alvo, de modo que se possa obter previsões mais confiáveis. / The success of opinion mining for automatically processing vast amounts of opinionated content available on the Internet has been demonstrated as a less expensive and lower latency solution for gathering public opinion. In this work, we investigate whether it is possible to predict variations in vote intention based on sentiment time series extracted from news comments, using three Brazilian elections as case study. The contributions of this case study are: a) the comparison of two approaches for opinion mining in user-generated content in Brazilian Portuguese; b) the proposition of two types of features to represent sentiment behavior towards political candidates that can be used for prediction, c) an approach to predict polls vote intention variations that is adequate for scenarios of sparse data. We developed experiments to assess the influence on the forecasting accuracy of the proposed features, and their respective preparation. Our results display an accuracy of 70% in predicting positive and negative variations. These are important contributions towards a more general framework that is able to blend opinions from several different sources to find representativeness of the target population, and make more reliable predictions.

User Generated Content (UGC) blir allt vanligare på webben och flera av de mest besökta hemsidorna på internet är till stor del baserade på innehåll skapat av dess användare. Varje dag redigeras exempelvis över 120,000 artiklar på Wikipedia av dess hängivna användare som i snitt spenderar över 8 timmar i veckan på att skapa nytt eller redigera existerande material. Att motivera användare att bidra med innehåll på detta sätt blir allt viktigare. För att kunna göra detta krävs en förståelse för vad som motiverar användarna att bidra. Uppsatsen tar sin utgångspunkt från detta och vi studerar vilka motivationsfaktorer som ligger till grund för att användare skapar UGC samt om det finns några samband mellan dessa motivationsfaktorer. En litteraturstudie över tidigare forskning inom området genomförs och resulterar i 15 motivationsfaktorer kategoriserade i fyra kategorier. Dessa presenteras i en matris som strukturerar upp motivationsfaktorerna i de olika kategorierna. Matrisen med de 15 motivationsfaktorerna ligger som teoretisk grund för en kvantitativ enkätundersökning ämnad att se i vilken grad användare motiverats av olika motivationsfaktorer för skapandet av UGC. Genom enkätundersökningen etableras en förståelse för hur stor andel av respondenterna som motiverats av de olika motivationsfaktorerna. Resultatet från undersökningen används för att studera sambanden mellan motivationsfaktorerna. Detta leder fram till en korrelationstabell som presenterar alla signifikanta samband mellan faktorer samt en motivationsmatris som kan fungera som ett stöd för att förstå och diskutera kring motivationsfaktorer för UGC och dess kategoriseringar. Studiens resultat kan hjälpa utvecklare att ge stöd för flera motivationsfaktorer i tjänster baserade på UGC. Den motivationsmatris och det resultat som presenteras kan även vara utgångspunkt för framtida forskning. / User Generated Content (UGC) is becoming increasingly common on the web and many of the most frequently visited websites on the Internet is largely based on content created by its users. Every day 120,000 articles gets edited on Wikipedia by its devoted users who spend an average of eight hours a week on creating new or edit existing material. To motivate users to contribute with content is becoming increasingly important. This requires an increased understanding of what motivates users to contribute.  In this paper we study the motivational factors that motivates users to contribute with UGC and if there is any correlation between these motivational factors. A literature review of previous research in the field of UGC is performed and results in 15 motivational factors categorized into four categories. These are presented in a matrix that structures the motivation factors in the different categories. The matrix of the 15 motivational factors are the theoretical foundation for a quantitative survey designed to see to what extent users are motivated by different motivational factors for the creation of UGC. Through the survey we establish an understanding to what extent different motivational factors motivate the participiants of the study to create UGC. The results from the survey are then used to study the relationships between motivational factors. This leads to a correlation table that presents all the significant relationships between factors and a motivation matrix that can serve as a basis for understanding and discuss the motivators for the UGC and its categorizations. Our results may help developers to support multiple motivational factors in services based on UGC. The motivation matrix and the results presented may also be the basis for future research.