Harnessing Semantics for Answer Sentence Retrieval
Proceedings of the 8th Workshop on Exploiting Semantic Annotations for Information Retrieval (ESAIR'15), 2015
Finding answer passages from the Web is a challenging task that has yet been thoroughly studied. One major difficulty is to retrieve sentences that may not have many terms in common with the question. In this paper, we experiment with two semantic approaches for finding non-factoid answers using a learning-to-rank retrieval setting. We show that using semantic representations learned from external resources such as Wikipedia or Google News may substantially improve the quality of top-ranked retrieved answers.
@inproceedings{chen2015harnessing,
author = {Chen, Ruey-Cheng and Spina, Damiano and
Croft, W. Bruce and Sanderson, Mark and Scholer, Falk},
title = {Harnessing Semantics for Answer Sentence Retrieval},
booktitle = {Proceedings of the 8th Workshop on Exploiting Semantic Annotations
for Information Retrieval (ESAIR'15)},
year = {2015}
}