Learning Domain-Specific Knowledge from Context--THUIR at TREC 2005 Genomics Track

We(Tsinghua University) participated both Ad Hoc Retrieval Task and Categorization Task in TREC2005 Genomics Track, in which we designed and implemented a serious of methods encompassed learning domain-specific knowledge from context. In Ad Hoc Retrieval Task, internal resource is introduced to expand query, different granularity indexing provides more flexible retrieval space, and pattern discovering imports Information Extraction (IE) concept into Information Retrieval (IR). In Categorization Task, instead of the single word feature, we presented Seed-based Loose N-gram Feature, which achieved success in the four subtasks.