COMBINING THE EVIDENCE OF MULTIPLE QUERY REPRESENTATIONS FOR INFORMATION-RETRIEVAL

We report on two studies in the TREC-2 program that investigated the e ffect on retrieval performance of combination of multiple representati ons of TREC topics. In one of the projects, five separate Boolean quer ies for each of the 50 TREC routing topics and 25 of the TREC ad hoc t opics were generated by 75 experienced online searchers, Using the INQ UERY retrieval system, these queries were both combined into single qu eries, and used to produce five separate retrieval results for each to pic. In the former case, progressive combination of queries led to pro gressively improving retrieval performance, significantly better than that of single queries, and at least as good as the best individual si ngle-query formulations. In the latter case, data fusion of the ranked lists also led to performance better than that of any single list. In the second project, two automatically produced vector queries and thr ee versions of a manually produced P-norm extended Boolean query for e ach routing and ad hoc topic were compared and combined. This project investigated six different methods of combination of queries, and the combination of the same queries on different databases. As in the firs t project, progressive combination led to progressively improving resu lts, with the best results, on average, being achieved by combination through summing of retrieval status values. Both projects found that t he best method of combination often led to results that were better th an the best performing single query. The combined results from the two projects have also been combined by data fusion. The results of this procedure show that combining evidence from completely different syste ms also leads to performance improvement.