Perspectives for Rank Aggregation within Scientific Publication Databases

Ranking in scientific publication databases involves a variety of additional resources that are usually not applied in standard general purpose search engines. Moreover, community-specific expectations of users influence the perception of the adequacy of retrieved results. Architects of specialized search engines therefore face an additional challenge of integrating complementary knowledge into the information retrieval system to cope with the specific user needs. In this sense the notion of ranking is extended by the need to incorporate various ranking methods exploiting these resources providing a way to rank documents by individual criteria. Integrating these individual ranking methods into one aggregation scheme becomes then a central issue. In this perspective, as every specialized domain relies on a different range of ranking methods and resources, the targeted model needs to be generic in a sense that it must allow for incorporating the particularities of a given domain in a standardized way. In this document we present the perspectives of the rank aggregation within scientific publication databases based on our experience from the work on the CERN document Server collection of documents in the domain of particle and high energy physics.