Robust pseudo feedback estimation and HMM passage extraction: UIUC at TREC 2006 genomics track

Jing Jiang, Xin He, Chengxiang Zhai

Research output: Contribution to journalConference articlepeer-review

Abstract

In summary, in this year's Genomics Track, we focused on testing the effectiveness of two language modeling techniques for information retrieval on biomedical text. The general observation is that the two techniques are still effective to some degree on biomedical text. The regularized feedback estimation method is more robust than the original feedback estimation method because it needs less parameter tuning. The HMM-based passage extraction method can outperform paragraph-based passages. However, since the HMM-based method is not designed to extract short passages with very specific information, it needs some modification in order to fit this task. Finally, user relevance feedback is very effective.

Original languageEnglish (US)
JournalNIST Special Publication
StatePublished - Dec 1 2006
Event15th Text REtrieval Conference, TREC 2006 - Gaithersburg, MD, United States
Duration: Nov 14 2006Nov 17 2006

ASJC Scopus subject areas

  • Engineering(all)

Fingerprint Dive into the research topics of 'Robust pseudo feedback estimation and HMM passage extraction: UIUC at TREC 2006 genomics track'. Together they form a unique fingerprint.

Cite this