Applying Inference Networks for Multiple Collection Searching
Zhigang Lü · 1996
The paper describes how to use inference networks to solve two problems in searching multiple collections: collection selection and result merging. The effectiveness of the approaches is demonstrated with the INQUERY system and 3 gigabyte TREC collections. 1 Introduction As hundreds or even thousands of collections are available on the Internet, the IR community must cope with the problem of searching multiple collection. The simplest approach is to build a single index for all collections. But this approach is practically prohibited by its obvious drawback: it is too slow, because searching such a gigantic index takes a long time to complete. Worse, this search may not complete due to network resource limits, such as limited bandwidth in the case of hundreds of collections. One way to get a quick response is to narrow the search to a portion of the index, which leads to a partition of the index. The natural way to partition is for each partition to correspond to one collection. Heuri...