Facet-Driven Blog Feed Retrieval

Abstract

The faceted blog distillation task retrieves blogs (i.e. RSS feeds) that are not only relevant to a query but also satisfy an interested facet. The facets under consideration are opinionated vs. factual, personal vs. official and in-depth vs. shallow. For the opinionated/factual facets, we propose a classifier that uses syntactic and semantic features to determine whether an opinion in blog documents is relevant to a given query. For the personal/official facets, we propose three classifiers that are learned based on different assumptions to categorize a blog document into either the personal or the official class. For the in-depth/shallow facets, we propose to calculate the depth of the coverage of a blog document on a given query by the occurrences of the concepts related to the query. Dependencies among different facets are also discussed. Experimental results on TREC Blogs06 and Blogs08 collections show that our techniques are not only effective in finding faceted blogs but also significantly outperform the best known results over both collections.