Results Processing in Heterogeneous Digital Libraries

摘要

In the heterogeneous digital libraries, users are allowed to have access to data of different modalities, from different information sources, and ranked by different criteria. This paper,which assumes that each information retrieval model is satisfactory in its own context, proposes two results processing methods: Ranking by Sources (RBS) and Simply Merging Results (SMR). It defines satisfied ranking, which satisfies most source rankings, and satisfied distance to indicate how a specific source ranking suits the satisfied ranking. RBS ranks sources by their satisfied distances and groups the results by sources. SMR, using source scoring function distributions, substitutes normalized scores for original scores from sources, and then merges the results using these normalized scores. The results show they are very feasible and efficient in the heterogeneous