Coronavirus Update

All academic classes will be held remotely for summer semester and course information indicates whether they have set meet times or no meet times. Physical access to Dayton and Lake campuses is restricted to essential personnel. Read more.

Ph.D. Dissertation Defense “Data-driven and Knowledge-driven strategies for Realizing Crowd Wisdom on Social Media” By Shreyansh Bhatt

Thursday, June 13, 2019, 9:30 am to 11:30 am
366 Joshi Research Center
Current Students

Ph.D. Committee:  Drs. Amit Sheth, Advisor, Keke Chen, Krishnaprasad Thirunarayan, Valerie Shalin (Department of Psychology) and Brandon Minnery (Kairos Research)


The wisdom of the crowd is a well-known example of collective intelligence wherein an aggregated judgment of a group of individuals is superior to that of an individual. The aggregated judgment is surprisingly accurate to predict the outcome of a range of events from geopolitical forecasting to stock price index. Recent research studies have shown that participants’ previous performance data contributes to the identification of a subset of participants that can collectively predict an accurate outcome. In the absence of such performance data, researchers have explored the role of human-perceived diversity to assemble an intelligent crowd. The online social networks are becoming increasingly popular in sharing and seeking domain-specific knowledge. Tapping the crowd wisdom on online social networks can help prediction for several real-world tasks. Independent and contextually diverse crowd selection using social media data imposes unique challenges such as,

  • Complementing short and potentially noisy social media data with domain specific knowledge.
  • Lack of labeled data in evaluating closely connected social media participants.
  • Combining social media text, indicating a diverse perspective, and network features, indicating potential influence, in diverse crowd selection.
  • Interpretable diversity measure to understand the type of diversity that can enhance the performance of a crowd.

This dissertation first provides several data-driven measures from social-media data and shows that participant diversity can be inferred from social media data and that it can benefit performance in the real world prediction tasks. Domain-specific knowledge graphs provide the foundational basis to evaluate and drive contextually diverse crowd selection and complementing short social media text. Novel group detection and crowd selection algorithms incorporating text, network, and knowledge-graph can automatically select diverse crowd and also provide recognizable diversity interpretation. It is shown that such a diverse crowd can accurately predict an outcome of real-world events. These results have implications for numerous domains that utilize aggregated judgments - from consumer reviews to econometrics, to geopolitical forecasting and intelligence analysis.

For information, contact
Log in to submit a correction for this event (subject to moderation).