Svetlana S. Bodrunova

Interval Semi-Supervised LDA: Classifying Needles in a Haystack

An important text mining problem is to find, in a large collection of texts, documents related to specific topics and then discern further structure among the found texts. This problem is especially important for social sciences, where the purpose is to find the most representative documents for subsequent qualitative interpretation. To solve this problem, we propose an interval semi-supervised LDA approach, in which certain predefined sets of keywords (that define the topics researchers are interested in) are restricted to specific intervals of topic assignments. We present a case study on a Russian LiveJournal dataset aimed at ethnicity discourse analysis.

Comparing efficacy of web design of university websites: mixed methodology and first results for Russia and the USA

Understanding the mechanisms of visual perception is important in the context of both media research and its applications in design practice. Within the functional approach to interface design, eye tracking is an established method to analyze interface efficacy. At the same time, in todays media design, many rules have been established by practitioners and remain untested. In this mixed-method study, we combine web crawling, web analytics and heat map analysis based on eye tracking, and qualitative usability analysis of composite-graphic model of a website. We check whether eye tracking test results (numeric data and heat map analysis) correlate to usability of key pages of a large website, as measured qualitatively according to recommendations of leading design literature. Among large web spaces, university website clusters represent a special type and suit well for our analysis, as they unite very different publics and are multi-task. We elaborate and pre-test the methodology on three sites of leading universities in the USA and Russia (Harvard University, Moscow State University and St.Petersburg State University). Our results suggest that there is no direct link between design-based elements of page usability and numeric eye tracking data, but heat maps show correlation with design quality; this means we need to continue checking the suggested methodology on larger number of assessors.

Please Follow Us: Media roles in Twitter discussions in the United States, Germany, France, and Russia

The media are normatively expected to play significant roles in conflictual discussions within national and international communities. As previous research shows, digital platforms make scholars rethink these roles based on media behavior in online communicative environments as well as on the structural limitations of the platforms. At the same time, traditional dichotomies between information dissemination and opinion formation roles, although seemingly universal, also vary across cultures. We look at four recent conflicts of comparable nature in the United States, Germany, France, and Russia to assess the roles that legacy media have performed in the respective ad hoc discussions on Twitter. Our approach differs from previous studies, as we combine content analysis of tweets by the media and journalists with the resulting positions of the media in the discussion graphs. Our findings show that, despite the overall trend of the “elite” and regional media sticking to information dissemination, online-only media and individual journalists vary greatly in their normative strategies, and this is true across countries. We also show that combining performance in content and social network analysis may allow for reconceptualization of media roles in a more flexible way.

U-index: An Eye-Tracking-Tested Checklist on Webpage Aesthetics for University Web Spaces in Russia and the USA

Background. Understanding the relations between user perception and aesthetics is crucial for web design. But it is frequent in today’s graphic and media design that rules, established by practitioners even before the advent of Internet and still untested empirically, are taught at design schools and widely used for online interface design. So far, there is no well-established linkage between the in-class recommendations and our empirical knowledge on usability, for which design plays a role just as crucial as web projecting. Will webpages that are better from the designers’ viewpoint perform better in terms of usability? And can one have a list of recommendations tested empirically?

Stable Topic Modeling with Local Density Regularization

Topic modeling has emerged over the last decade as a powerful tool for analyzing large text corpora, including Web-based user-generated texts. Topic stability, however, remains a concern: topic models have a very complex optimization landscape with many local maxima, and even different runs of the same model yield very different topics. Aiming to add stability to topic modeling, we propose an approach to topic modeling based on local density regularization, where words in a local context window of a given word have higher probabilities to get the same topic as that word. We compare several models with local density regularizers and show how they can improve topic stability while remaining on par with classical models in terms of quality metrics.

Multi-dimensional Echo Chambers: Language and Sentiment Structure of Twitter Discussions on the Charlie Hebdo Case

Background. Public discussions on social networks have trans-border and multilingual nature. This is especially true for conflictual discussions that reach global trending topics. Being part of the global public sphere, such discussions were expected by many observers to become horizontal, all-involving, and democratically efficient. But, with time, criticism towards the democratic quality of discussions in social media arose, with many works discovering the patterns of echo chambering in social networks. Even if so, there is still scarce knowledge on how affective hashtags work in terms of user clusterization, as well as on the differences between emotionally ‘positive’ and ‘negative’ hashtags. Objectives. We address this gap by analyzing the Twitter discussion on the Charlie Hebdo massacre of 2015. In this discussion, the Twittershpere has created #jesuischarlie and #jenesuispascharlie - two discussion clusters with, allegedly, opposite sentiments towards the journal’s ethics and freedom of speech. Research design. We were interested in whether echo chambers formed both on the hashtag level (based on language use) and within a language (based on user sentiment of French-speaking users). For data collection, we used vocabulary-based Twitter crawling. For data analysis, we employed network analytics, manual coding, web graph reconstruction, and automated sentiment analysis. Results. Our results show that #jesuischarlie and #jenesuispascharlie are alike in language distribution, with French and English being the dominant languages and the discussions remaining within the Euro-Atlantic zone. The language-based echo chambers formed in both cases. But if #jesuiuscharlie was a clear sentiment crossroads, #jenesuispascharlie was a negative echo chamber, thus allowing us to draw conclusions about multi-layer echo chambering.

Impact of Menu Complexity upon User Behavior and Satisfaction in Information Search

Background. The growing complexity of website navigation demands more behavior-oriented research. Today, static and sequential menus have become virtually incompatible with adaptive forms of web layouts; tagging-based menus started to dominate; and these navigational elements more and more co-exist on webpages. Previous research compares the three basic types of menus (static, sequential, and expandable) for two dimensions: reduction of mental load (objective, linked to task complexity and structural complexity) and growth of user satisfaction (subjective, linked to menu type). Objectives. We hypothesized that growth of task complexity linked to menu complexity leads to selection of non-productive search strategies and to growth of perceived complexity of the interface. Research design. Following Kang et al. (2008), we have divided user search strategies into productive (systemic) and non-productive (chaotic) and have conducted an experimental pre-test. Menu complexity was created by using four different menus in one prototype. Structural complexity was assessed by path depth and menu options diversity. The search tasks were designed to be realizable disregarding the menu complexity. Two homogenous groups of 10 assessors were consecutively conducting tasks on six HTML pages of the prototype. A questionnaire was used to assess user satisfaction. Results. For low-complexity tasks, menu diversity has virtually no impact upon navigational behavior. But we have discovered impact of menu complexity for high-complexity tasks for multi-menu navigational schemes. Additional tests with newer types of menus show that they make the assessors drop the sequential principle of search. Also, these pages were perceived as the hardest to use.

Comparing Influencers: Activity vs. Connectivity Measures in Defining Key Actors in Twitter Ad Hoc Discussions on Migrants in Germany and Russia

Today, a range of research approaches is used to define the so-called influencers in discussions in social media, and one can trace both conceptual and methodological differences in how influencers are defined and tracked. We distinguish between ‘marketing’ and ‘deliberative’ conceptualization of influencers and between metrics based on absolute figures and those from social network analytics; combining them leads to better understanding of user activity and connectivity measures in defining influential users. We add to the existing research by asking whether user activity necessarily leads to better connectivity and by what metrics in online ad hoc discussions, and try to compare the structure of influencers. To do this, we use comparable outbursts of discussions on inter-ethnic conflicts related to immigration. We collect Twitter data on violent conflicts between host and re-settled groups in Russia and Germany and look at top20 user lists by eight parameters of activity and connectivity to assess the structure of influencers in terms of pro/contra-migrant cleavages and institutional belonging. Our results show that, in both discussions, the number of users involved matters most for becoming an influencer by betweenness and pagerank centralities. Also, contrary to expectations, Russian top users all in all are, in general, more neutral, while Germans are more divided, but in both countries pro-migrant media oppose anti-migrant informal leaders.

Content Sharing in Conflictual Ad - Hoc Twitter Discussions: National Patterns or Universal Trends?

Recently, the growing role of social network users in content dissemination has brought to life the concept of secondary gatekeeping – selection and republication of content already selected and published by traditional gatekeepers. Secondary gatekeeping is believed to be raising the media in-platform visibility, but it may also have negative effects such as adding to creation of echo chambers and deepening the gaps between conflicting views. Such studies are particularly relevant for emergencies or social conflicts where sharing relevant content may be crucial for lowering social unease. But till today the nature of secondary gatekeeping remains highly understudied. We have conducted a comparative study of three ad-hoc Twitter discussions on heated ethnic/racial conflicts in the USA (Ferguson riots), Germany (Koln mass abuse), and Russia (Biryulyovo anti-migrant bashings) to assess the patterns of content sharing by active discussants. We used vocabulary-based web crawling and human coding of over 1,000 tweets in randomized samples. Our results show that, in all cases, there’s weak but significant correlation between the type of user and his/her attitude to minority with the attitudes expressed in content, while it is not always true that users prefer the same gatekeeper type, e.g. online or social media. As difference between individual users remains statistically significant, this may mean that the nature of heated ad-hoc discussions facilitates formation of ‘individual-level filter bubbles’ in addition to bigger echo chambers.

Ad-hoc publics and influencers on the Russian Twitter: The case of discussion on migrants

With the emergence of discussion platforms like Twitter, the hopes rose that computermediated public sphere would become more even in access to discussion than massmediatized public sphere of the late 20 century. Scholars have argued that it will eventually form an ‘opinion crossroads’ where conflicts would be discussed by all the parties involved. But today, existing research provides mixed evidence on whether ordinary users, rather than mainstream media and institutional actors, can become influencers in discussions on current issues, e.g. relations between host and migrant communities. We focus on the Twitter discussion about an inter-ethnic conflict in Moscow’s Biryuliovo district in 2013, as well as the comparative ‘calm’ period in March 2014, and look at who were real influencers by reconstructing the discussion’s web graph, as well as analyzing and juxtaposing its metrics to figures indicating user activity. Our results show that ad hoc discussion differs dramatically from an issuebased one in terms of the influencer nature and composition; the role of active tweeting is questioned. We also show that nationalist accounts play a much bigger role than expected in both periods.