A Position-Aware Deep Model for Relevance Matching in Information Retrieval.
Kai-Lung Hui, Andrew Yates, Klaus Berberich, Gerard de Melo · arXiv (Cornell University) · 2017
In order to adopt deep learning for information retrieval, models are needed that can capture all relevant information required to assess the relevance of a document to a given user query. While previous works have successfully captured unigram term matches, how to fully employ position-dependent information such as proximity and term dependencies has been insufficiently explored. In this work, we propose a novel neural IR model named PACRR (Position-Aware Convolutional-Recurrent Relevance), aiming at better modeling position-dependent interactions between a query and a document via convolutional layers as well as recurrent layers. Extensive experiments on six years' TREC Web Track data confirm that the proposed model yields better results under different benchmarks.