UIC at TREC 2010 Faceted Blog Distillation
Lifeng Jia, Clement T. Yu · 2010
Our system consists of a concept-based retrieval subsystem which performs the baseline blog distillation, an opinion identification subsystem and an opinion-in-depth analysis subsystem which performs the faceted blog distillation task. In the baseline task, documents which are deemed relevant are retrieved by the retrieval system with respect to the query, without taking into consideration of any facet requirements. The feeds are ranked in descending order of the sum of the relevance scores of retrieved documents. In order to improve the recall of the retrieval subsystem, we recognize proper nouns or dictionary phrases without requiring matching all the words of the phrases. In the opinionated vs. factual and personal vs. official faceted tasks, the opinion identification subsystem is employed to recognize query-relevant opinions within the documents. Personal documents are more likely to be opinionated than official documents. In the in-depth vs. shallow faceted task, the depth of the opinion within a document is measured by the number of concepts which are related with the query the document contains. 1.