This research project is an effort to integrate domain knowledge of a special cultural interest with novel search capabilities for assessing political risks. The key technical issues of this project include: (i) how to start the search wisely; (ii) how to search smartly to obtain sensible results in reasonable amount of time; (iii) how to refine the search with increased knowledge via analysis of searchable results and explore the vast blogosphere efficiently beyond the known; and (iv) how to judiciously track groups for studying their evolution. We develop algorithms for searching the long tail to connect bloggers of similar interest. A significant achievement is that we propose and experiment a collective-wisdom based search approach that paves the way of efficiently connecting sparsely distributed similar bloggers. A preliminary system has taken its shape in September 2008 with minimum functions for demonstration purposes. The first phase of the proposed project is planned for the period of January 29 and November 30, 2008. An annual report for the project ending May 31st, 2008 was submitted with specific achievements (publications, awards, etc.) The second phase of the project consists of two major parts: the first part is about crawling and storage of blog data streams, and the second part is about interactive presentation, comparative analysis, and visualization. In the second phase of project, we propose to conduct research with four interrelated research threads: (1) continuing algorithmic research on efficient search of similar bloggers; (2) developing functions of BlogTrackers designed for monitoring of multiple blogs of interest; (3) an interactive environment for experimentation and validation that assists and takes input from a user (e.g., a non-computer scientist but domain expert); and (4) education and dissemination of research findings.
|Effective start/end date||10/10/08 → 5/31/10|
- DOD-NAVY: Office of Naval Research (ONR): $88,981.00