Small Talks about BIG DATA

small talks about BIG DATA

Hosted by the Business Intelligence & Analytics program / faculty

June 12, 2012: Winter Mason, Assistant Professor

The Emergence of Social Conventions in Online Social Networks

Although social conventions are a powerful guide for behavior, the way they emerge in communities is not well understood. We focus on competing conventions for attributing reposts to the original source on Twitter. We analyze over 1.7 billion "tweets" from 54 million users, and observe how the conventions emerged and spread through the network of Twitter users. Initially the most successful conventions were borrowed from natural language ("via" and "retweeting"), but over time a community-specific convention came to dominate ("RT"). Additional evidence suggests this specificity to the community, along with efficiency of communication and the timing of the adoption, are key variables in the acceptance of the convention. We build models that suggest there are some features that encourage the adoption of one convention over another, but that there remains significant inherent unpredictability in what convention will ultimately dominate.

Winter Mason is an Assistant Professor of Information Systems in the Howe School of Technology.  He received his Bachelor’s of Science degree in Psychology from the University of Pittsburgh in 1999, and his Ph.D. in Social Psychology and Cognitive Science in 2007 from Indiana University.  From 2007–2011, he was a Visiting Scientist at Yahoo! Research in the Human and Social Dynamics group in New York City, at which point he joined Stevens Institute of Technology.  His research is focused on crowdsourcing, social networks, and group dynamics. Methodologically, his research spans traditional psychological methods including laboratory experiments, new methods such as online data collection with crowdsourcing, and computer science methods such as data mining.

June 13, 2012:  Germán Creamer, Associate Professor

Can Corporate News Network Influence Volatility and Prices?

  - joint work with Yong Ren, & Jeffrey V. Nickerson

We evaluate the impact of corporate news network in the price dynamic of European stocks for the period 2005-2011. We performed several non-linear tests of independence among several social network indicators, prices and volatility. The causality tests suggest it is possible to improve the prediction of prices and volatility by extracting and analyzing a network based on the mentions of companies in news stories.

Germán Creamer is an Associate Professor of quantitative finance and financial engineering at Stevens Institute of Technology. Dr. Creamer has been a senior manager in the Risk, Information and Banking Division in American Express where he worked in the enterprise-wide risk management and the information management groups. He has also taught at Columbia University, Tulane University, and in several leading Latin American business schools.

Howe.School@stevens.edu for more information