Counting on TION

Most test collections (like TREC and CLEF) for experimental research in information retrieval apply binary relevance assessments. This paper introduces a four-point relevance scale and reports the findings of a project in which TREC-7 and TREC8 document pools on 38 topics were reassessed. The goal of the reassessment was to build a subcollection of TREC for experiments on highly relevant documents and to learn about the assessment process as well as the characteristics of a multigraded relevance corpus. Relevance criteria were defined so that a distinction was made between documents rich in topical information (relevant and highly relevant documents) and poor in topical information (marginally relevant documents). It turned out that about 50% of documents assessed as relevant were regarded as marginal. The characteristics of the relevance corpus and lessons learned from the reassessment project are discussed. The need to develop more elaborated relevance assessment schemes is emphasized.