Name
Abstract | ||
|---|---|---|
This paper describes the techniques we applied for the two tasks of the TREC Genomics track, i.e., ad hoc retrieval and categorization tasks. For the ad hoc retrieval task, we used query expansion, different scoring strategy on different parts of Medline record (Title, Abstract, RN, MH, etc.) and pseudo relevance feedback. Our submitted run DUTAdHoc2 obtained a MAP of 0.2349. For the categorization task, our system used a SVM classifier with TFIDF term weighting scheme. In addition concept replacing and filtering methods were adopted. Two of our submitted runs (eDUTCat1 and gDUTCat1) produced a Utility score of 0.8496 and 0.572 respectively ranking third and fourth out of 46 runs submitted for the categorization task. |
Year | Venue | Keywords |
|---|---|---|
2005 | TREC | query expansion |
Field | DocType | Citations |
Data mining,Categorization,Weighting,Relevance feedback,Query expansion,tf–idf,Information retrieval,Ranking,Computer science,Filter (signal processing),TREC Genomics | Conference | 2 |
PageRank | References | Authors |
0.38 | 4 | 5 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Zhihao Yang | 1 | 270 | 36.04 |
| Hongfei Lin | 2 | 768 | 122.52 |
| Yanpeng Li | 3 | 7 | 1.66 |
| Baoyan Liu | 4 | 2 | 1.39 |
| Ye Lu | 5 | 19 | 3.51 |