Passage Feedback for News Tracking
Abstract
We extend the event tracking task of Topic Detection and Tracking (TDT) to create a framework in which a user can highlight relevant passages in addition to specifying the relevance of documents. A dual framework of combined document and passage feedback improves performance over a state-of-the-art system without feedback by over 70%. Although annotators vary in the content and length of the passages marked for feedback, improvements in performance are consistent, We demonstrate how events in news follow certain trends over time, making passage feedback critical to event tracking.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2006
- Accession Number
- ADA449605
Entities
People
- Hema Raghavan
- James Allan
Organizations
- University of Massachusetts Amherst