Deep Neural Network for Learning to Rank Query-Text Pairs
This addresses document ranking for information retrieval systems, presenting an incremental improvement over existing methods.
The paper tackles document ranking in information retrieval by proposing ConvRankNet, which combines a Siamese Convolutional Neural Network encoder with the RankNet ranking model for end-to-end training. Experiments on the OHSUMED dataset show it systematically outperforms existing feature-based models.
This paper considers the problem of document ranking in information retrieval systems by Learning to Rank. We propose ConvRankNet combining a Siamese Convolutional Neural Network encoder and the RankNet ranking model which could be trained in an end-to-end fashion. We prove a general result justifying the linear test-time complexity of pairwise Learning to Rank approach. Experiments on the OHSUMED dataset show that ConvRankNet outperforms systematically existing feature-based models.