Sikana Tanupabrungsun

Understanding Discourse Acts: Political Campaign Messages Classification on Facebook and Twitter

To understand political campaign messages in depth, we developed automated classification models for classifying categories of political campaign Twitter and Facebook messages, such as calls-to-action and persuasive messages. We used 2014 U.S. governor’s campaign social media messages to develop models, then tested these models on a randomly selected 2016 U.S. presidential campaign social media dataset. Our classifiers reach .75 micro-averaged F value on training sets and .76 micro-averaged F value on test sets, suggesting that the models can be applied to classify English-language political campaign social media messages. Our study also suggests that features afforded by social media help improve classification performance in social media documents.

Call to Retweet: Negotiated Diffusion of Strategic Political Messages

Twitter allows political candidates to broadcast messages directly to the public, some of which spread virally and potentially reach new audiences and supporters. During the 2014 U.S. gubernatorial election, 74 candidates posted 20,580 tweets, of which, 10,946 were retweeted a total of 139,315 times. Using content analysis, automated classification and regression analysis, we show that actors with different levels of network influence tend to promote different types of election content, but that the convergence of their choices and actions lead to information flows that reach the largest audiences. We also show that actors with middle-level influence, in terms of the number of followers they have, tend to be the most influential in the diffusion process. Our work provides empirical support for the theoretical framework of negotiated diffusion, which suggests that information flows are the result of the convergence of top-down forces (structures and powerful gatekeepers) and bottom-up forces (collective sharing of actors with varying degrees of influence).

Noisy candidates and informative politicians: analyzing changes in tweet behavior using tweet quality assessment framework

Politicians use Twitter as a strategic tool for campaigning and posting messages once elected. Our work focuses on the ways U.S. state governors’ use twitter differently when they were campaigning vs. after they have taken office. Our data consists of tweets posted by wining gubernatorial candidates during and six months after the 2014 elections. Using regression analysis, we find that post-election tweet volume is related to factors such as pre-election tweet volume, incumbency status, and if they were a third party candidate. We also develop and utilize a novel Tweet Quality Assessment Framework (TQAF) to show that during elections politicians try to engage topically with their audience more than once elected, but tend to produce higher quality information once in office. Our work contributes to the understanding of politician’s use of Twitter. We also believe our TQAF will be useful for researchers wishing to compare differences in tweet behavior across time or groups.

Studying Celebrity Practices on Twitter Using a Framework for Measuring Media Richness

Social media enables the performative actions needed for celebrities to build and maintain audiences. Platforms like Twitter mediate identity construction and interaction with fans while enabling environments that are co-constructed by celebrities, fans, and the platform itself. We use the theoretical lens of media richness to study the ways that different types of celebrities enact “micro-celebrity” by mapping three richness dimensions (contextual, interactional, and informational) into groupings of Twitter’s affordances. Utilizing crowdsourcing and regression analysis, we systematically weigh each affordance and generate richness scores for each dimension, for each tweet. Using these richness scores, we find that performance of different types of celebrity requires different affordance mixtures, and that these mixtures reflect differences in the environments within which celebrities operate. Our research contributes to work at the intersection of Twitter affordances and celebrity studies in new media, and provides a framework for generating richness scores.

Social Media, Opinion Polls, and the Use of Persuasive Messages During the 2016 US Election Primaries:

Political campaigns’ use of digital technologies has been a topic of scholarly concern for over two decades, but most studies have been focused on analyzing the use of digital platforms without considering contextual factors of the race, like public opinion polls. Opinion polls are an important information source for citizens and candidates and provide the latter with information that might drive strategic communication. In this article, we explore the relationship between the use of social media in the 2016 US presidential elections and candidates’ standing in public opinion polls, focusing on the surfacing and primary stages of the campaign. We use automated content analysis to categorize social media posts from all 21 Republican and Democratic candidates. Results indicate that a candidate’s performance in the polls drives certain communicative strategies, such as the use of messages of attacks and advocacy, as well as the focus on personal image.

Viral Design: User Concepts of Virality on the Niche Social Media Site, Dribbble

Virality is a much-studied topic on popular social media sites, but has been rarely explored on niche sites. Dribbble is a niche social networking site for artists and designers with over 600,000 users. Using a mixed-method approach, we explore virality from a user-centric perspective. Interviews confirm that viral-like events do exist on Dribbble. Through interviews we identify the measures and possible driving factors of viral-like events. While what spreads is different than on other platforms, our work suggests that the measures and mechanics that drive these events are similar. These similarities reflect fundamental human behavior underlying social phenomenon across different platforms. Our results are supported by regression modeling using variables identified by our informants. Smaller sites like Dribbble are rarely studied, so our work contributes to social media studies, particularly using mixed methods approaches, and to the body of research around information diffusion and viral events.

Tweeting to the Target: Candidates’ Use of Strategic Messages and @Mentions on Twitter

ABSTRACT This paper reports on a mixed-methods (i.e., content analysis, machine learning, and quantitative analysis) study of Twitter use among 74 U.S. gubernatorial candidates during the 2014 election. In extending the theory of controlled interactivity, this article focuses on politicians’ use of the @mention where we detail differing messaging strategies when candidates mention themselves versus their opponents, and between incumbents and challengers. Results suggest that candidates use the @mention feature as a subtle audience targeting mechanism. Our work also offers a methodological contribution by showing that machine-learning models perform better when context variables are included.

Dribbble: Exploring the Concept of Viral Events on an Art World Social Network Site

While virality is a much-studied topic on popular social media sites, it has been rarely explored on sites like Dribbble, a social networking site for artists and designers. Using a mixed-method approach, we explore virality from a user-centric perspective. Interviews with informants confirm that viral-like events do exist on Dribbble, though what spreads are stylistic choices. While what spreads is different than on other platforms, our work suggests that the mechanics that drive these events are similar, suggesting an underlying social phenomenon that is reflected in different ways on different platforms. Our results are supported by regression modeling using variables identified by our informants. Our work contributes to social media studies since smaller sites like Dribbble are rarely studied, particularly using mixed methods approaches, as well as to the body of research around information diffusion and viral events.

Automated Diffusion? Bots and Their Influence During the 2016 U.S. Presidential Election

In the 2016 U.S. Presidential election, some candidates used to automated accounts, or bots, to boost their social media presence and followership. Categorizing all automated accounts as “bots” obfuscates the role different types of bots play in the spread of political information in election campaigns. Exploring strategies for automated information diffusion helps scholars understand and model online political behavior. This paper presents an initial effort aimed at understanding the disparate roles of bots in diffusion of political messages on Twitter. Having collected over 300 million tweets from candidates and the public from the U.S. presidential election, we use three OLS regression models to explore the strategic advantages of different types of automated accounts. We approach this by analyzing retweet events, testing a series of hypotheses regarding bots’ influence on the size of retweet events, and the change in candidates’ followers. Next, we develop an estimator to analyze the spread of information across the networks, demonstrating that, while ‘benevolent bots’ serve as overt information aggregators and have an effect on information diffusion, “nefarious bots” act as false amplifiers, covertly mimicking the spread of online information with no effect on diffusion. Making this important distinction allows us to disambiguate the concept of “bots” and reach a more nuanced and detailed understanding of the role of automated accounts in information diffusion in political campaigning online.

The text and images of the gopdebate: what the public is "talking" about on instagram

The richness of Instagram data makes it possible to tell a compelling story about the public’s “talk” on Instagram. Our work focuses on the use of Instagram by citizens to express their thoughts on the 2016 GOP presidential election. We collected Instagram posts with the hashtags #GOPDebate and #GOPDebates for five months. Using topic-modeling and sentiment analysis techniques to analyze both textual (post captions) and visual (images) attributes we are able to illustrate the topical network of political discussions and actors whom the public talks about on each topic. Our work contributes to literature on the role of social media, and specifically Instagram, in the political domain. The methodology also demonstrates how textual and visual attributes can be used together to categorize photo content.