Vortragende‎ > ‎

Peter Gloor - Abstract

Visual Discovery of Trends and Trendsetters Through Dynamic Semantic Social Network Analysis

Peter A. Gloor (MIT Center for Collective Intelligence)

Condor visualizes and analyzes communication activity, content and unstructured text, showing the evolution over time, while respecting individual privacy (Gloor & Zhao 2006). Condor has been developed for the last six years at MIT, Dartmouth College, University of Cologne, and others. It takes as input any type of communication archives such as e-mail, phone records, blogs, Weblinks, or f2f interaction records. An interactive movie shows the evolution of social networks over time, displays active relationships in a sliding time window, and calculates and plots the evolution of network metrics over time to discover interesting events in the lifetime of a virtual team and different phases in the life cycle of virtual communities. In addition to the social network view, the same algorithms are also applied to textual content analysis. Free text documents are parsed, and a keyword document frequency index is calculated.
The main advantage of Condor compared to other tools such as Ucinet, Pajek, Inflow, etc. is its capability to generate temporal social network movies combined with textual analysis.

Gloor, P. Zhao, Y. Analyzing Actors and Their Discussion Topics by Semantic Social Network Analysis, Proceedings of 10th IEEE International Conference on Information Visualisation IV06, London, 5-7 July 2006.

Web site: http://www.ickn.org


The right side of figure 1 illustrates Condor’s temporal elements. The contribution index view positions people based on the number of messages they exchanged, and terms based on the number of new documents in which they appear. The centrality view plots a continuous curve of the changes in group betweenness and degree centrality as well as graph density of both social network view and concept view.

Condor Movie about Migros Web Trends on YouTube

Condor Nanotech Blog Discourse on YouTube

Peter A. Gloor is a Research Scientist at the Center for Collective Intelligence at MIT's Sloan School of Management where he leads a project exploring Collaborative Innovation Networks and Chief Creative Officer of software startup galaxyadvisors.   He also lectures at the University of Cologne and Helsinki University of Technology. He has been working on visualizing knowledge for the last 15 years, leading to the development of Condor, a tool combining Web navigation with social network analysis. Previously, Peter was a Partner with Deloitte Consulting, a Partner with PricewaterhouseCoopers and a manager at UBS. pgloor@mit.edu