Self-disclosure topic model for Twitter conversations

Abstract

Self-disclosure, the act of revealing oneself to others, is an important social behavior that contributes positively to intimacy and social support from others. It is a natural behavior, and social scientists have carried out numerous quantitative analyses of it through manual tagging and survey questionnaires. Recently, the flood of data from online social networks (OSN) offers a practical way to observe and analyze self-disclosure behavior at an unprecedented scale. The challenge with such analysis is that OSN data come with no annotations, and it would be impossible to manually annotate the data for a quantitative analysis of self-disclosure. As a solution, we propose a semi-supervised machine learning approach, using a variant of latent Dirichlet allocation for automatically classifying self-disclosure in a massive dataset of Twitter conversations. For measuring the accuracy of our model, we manually annotate a small subset of our dataset, and we show that our model shows significantly higher accuracy and F-measure than various other methods.With the results our model, we uncover a positive and significant relationship between self-disclosure and online conversation frequency over time.

Cursing in English on twitter

“…Buzz communities and found that different communities (e.g., politics or sports) use profanity with varied frequencies and in different ways. Turning to Twitter, Bak et al [2] studied self-disclosure behavior in Twitter conversations. As one aspect of self-disclosure, profanity was used more frequently between users with higher relationship strength.…”

Self-disclosure topic model for classifying and analyzing Twitter conversations

“…We map each SD level to one sentiment and use the same seed words/trigrams from SECRET as in SDTM below. Used in previous work (Bak et al, 2012).…”

“…However, it is hard to identify intimacy between users in large scale online social network. So we choose conversation behaviors such as conversation frequency and length which can be treated as proxies for measuring intimacy between two people (Emmers -Sommer, 2004;Bak et al, 2012).…”

“…These methods consume much time and effort, so they are not suitable for large-scale studies. In prior work closest to ours, Bak et al (2012) showed that a topic model can be used to identify self-disclosure, but that work applies a two-step process in which a basic topic model is first applied to find the topics, and then the topics are post-processed for binary classification of self-disclosure. We improve upon this work by applying a single unified model of topics and self-disclosure for high accuracy in classifying the three levels of self-disclosure.…”

Microblog Analysis as a Program of Work

“…Examples of this kind of focus on Twitter abound. Bak et al [2012] looked at how certain kinds of content in Twitter conversations might be indicative of self-disclosure and the strength of relationship between the participants; Murnane & Counts [2014] examine the Twitter posts from smokers and the part they play in them giving up smoking; Kendall et al [2011] examine how people use Twitter to share information with one another about their health and fitness activities; Magee et al [2013] use tweets amongst players of a role-playing game to explore the part played by these in enhancing their gaming experience; Zhang et al [2011] look at how consumers use 'word-of-mouth' on Twitter to share information about businesses and their products and services; and Sleeper et al [2013] examine a range of tweets that tweeters have subsequently regretted posting and the actions they have then taken to engage in some kind of repair.…”