Learning to retrieve opinions

Abstract

As a novel information retrieval task, opinion retrieval has attracted considerable amount of attention in recent years. Current researches mainly first computed the topic relevant and opinion relevant scores of the documents and then combined these two scores as the final ranking score using some combination function. One major problem in existing works is that the score combination functions are defined in advance regardless of domains. However, there is no evidence that these two scores should be combined in a unique way. In this paper, we propose to learn the combination functions automatically for retrieval tasks of different domains. We employ the popular Genetic Programming framework for the learning tasks. To perform the whole opinion retrieval task, we also design a novel opinion retrieval system to compute the topic and opinion relevant scores and then learn the optimal combination function to integrate the topic and opinion relevant scores. In the experiments, we compare our system with other state-of-the-art work on a public dataset and the experimental results show that our system performs comparatively with others. © 2009 Springer-Verlag Berlin Heidelberg.