摘要:In this paper, we propose a novel approach for document re-ranking, which relies on the concept of negative feedback represented by irrelevant documents. In a previous paper, a pseudo-relevance feedback method is introduced using an absorbing document ~d which best fits the user's need. The document ~d is orthogonal to the majority of irrelevant documents. In this paper, this document is used to re-rank the initial set of ranked documents in Ad-hoc retrieval. The evaluation carried out on a standard document collection shows the effectiveness of the proposed approach.
关键词:Re-ranking; absorption of irrelevance; vector product