A Motif Approach for Identifying Pursuits of Power in Social Discourse

Abstract

In this paper, we investigate whether the social goals of an individual's utterances can be recognized through analysis of a discourse's intentional structure. Specifically we focus on identifying individuals pursuing power within a group. Individuals pursue power in order to increase their control of the actions and goals of the group. Following work in discourse processing we decompose the problem into identifying the social intention of the discourse segments and the intentional structure of the overall discourse. The set of social intentions come from eight psychologically-motivated social acts. We then build a motif-based representation of the discourse's social intentional structure that captures the interactions of the intentions between discourse participants. Finally we show how these structures can be used to identify the social goal of pursuit of power. Our best results achieve an accuracy of 84.2% for predicting pursuit of power in discussions communicated in English and 80.6% for discussions communicated in Chinese. I. INTRODUCTIONUnderstanding discourse in the modern world of social media presents a plethora of new challenges. One central challenge for the community is advancing our understanding of the semantics and pragmatics of social media. Communication over social media often has more to do with the social implications of the utterance than its content. However, current computational approaches to understanding the semantics and pragmatics of discourse focus only on identifying the actions of individuals towards the dialogue through dialogue acts [1], [2], and examination of the information content of the dialogue through topical analysis [3], [4]. Instead, in this paper we focus on identifying the social pragmatics of a discourse and present a computational model of the social goals of discourse participants. In particular, we focus on identifying individuals that are pursuing power within an online community.As in the many discourse theories that converged on the importance of discourse cues for parsing the discourse structure of dialogues [5], [6], we claim that knowledge about social cues and their expression in language is central for recognizing social acts in communication and intentions in discourse. Recent advances in syntactic, semantic, and discourse processing allow us to contemplate the automatic processing of sociolinguistic-driven properties by identifying the language uses that capture the manifestations of social acts within discourse segments. As in other textual inference tasks, e.g. textual entailment [7] and conversation entailment [8], there are a variety of syntactic and semantic constructs that are helpful in recognizing the social implications of a dialogue.

Computational Sociolinguistics: A Survey

“…Other studies have focused on a more dynamic view of power as arising through asymmetries with respect to needed resources or other goals, as characterized in consent-based theories of power such as exchange theory (Guinote and Vescio 2010). This would include such investigations as identifying persons who are pursuing power (Bracewell, Tomlinson, and Wang 2012;Swayamdipta and Rambow 2012) and detecting influencers (Biran et al 2012;Huffaker 2010;Nguyen et al 2014;Quercia et al 2011). This could also include studying how language use changes when users change their status in online communities (Danescu-Niculescu-Mizil et al 2012).…”

“…Many computational approaches within this sphere build on a foundation from pragmatics related to speech act theory (Austin 1975;Searle 1969), which has most commonly been represented in what are typically referred to as conversation, dialog or social acts (Bender et al 2011;Ferschke, Gurevych, and Chebotar 2012). Such categories can also be combined into sequences (Bracewell, Tomlinson, and Wang 2012). Other specialized representations are also used, such as features related to turn taking style (Prabhakaran, John, and Seligmann 2013;Swayamdipta and Rambow 2012), topic control (Nguyen et al 2014;Prabhakaran, Arora, and Rambow 2014;Strzalkowski et al 2012), and 'overt displays of power', which Prabhakaran, Rambow, and Diab (2012a) define as utterances that constrain the addressee's actions beyond what the underlying dialog act imposes.…”

Determining leadership in contentious discussions

“…Catizone [7] uses social network analysis methods to infer social and instrumental roles and relationships in online discussions. The work that comes closest to ours is by Bracewell [8], who define psychologically motivated acts to determine social goals of individuals based on discourses' intentional structure. Apart from network analysis methods, all the methods discussed here use annotators to characterize user behavior in the social network or discussion forums.…”

Technology Support for Discussion Based Learning: From Computer Supported Collaborative Learning to the Future of Massive Open Online Courses

“…Other work has its foundations in sociolinguistic analyses of dialogue structure (Prabhakaran et al 2012;Mayfield and Rosé 2011), Rhetorical Structure Theory (Allen et al 2014), the Pragmatics literature on Speech-Act theory (Cadilhac et al 2013;Ferschke et al 2012) or Politeness theory (Bramsen et al 2011), and Critical Discourse Analysis (Mukherjee et al 2013). In recent years, a substantial amount of work in this area has leveraged or at least references theoretical frameworks from Social Psychology (Bak et al 2014;Bracewell et al 2012). Within that sphere, some work has focused on more specific subcommunities such as the area of doctorpatient interaction (Wallace et al 2013) or the literature on social support and health (Mayfield et al 2013).…”