Relevance Propagation for Topic Distillation UIUC TREC 2003 Web Track Experiments

In this paper, we report our experiments on the Web Track TREC-2003. We submitted five runs for the topic distillation task. Our goal was to evaluate the standard language modeling algorithms for topic distillation, as well as to explore the impact of combining link and content information. We proposed a new general relevance propagation model for combining link and content information, and explored a number of specific methods derived from the model. The experiment results show that combining link and content information generally performs better than using only content information, though the amount of improvement is sensitive to the document collection and tuning of parameters.

• Type: Conference
• Year: 2003
• Where: TREC
• Authors: Azadeh Shakery, ChengXiang Zhai
• Keywords: Content Information, Topic Distillation, Topic Distillation Task, TREC 2003, TREC 2008