A comparison of genetic algorithms for optimizing linguistically informed IR in question answering

Abstract

In this paper we compare four selection strategies in evolutionary optimization of information retrieval (IR) in a question answering setting. The IR index has been augmented by linguistic features to improve the retrieval performance of potential answer passages using queries generated from natural language questions. We use a genetic algorithm to optimize the selection of features and their weights when querying the IR database. With our experiments, we can show that the genetic algorithm applied is robust to strategy changes used for selecting individuals. All experiments yield query settings with improved retrieval performance when applied to unseen data. However, we can observe significant runtime differences when applying the various selection approaches which should be considered when choosing one of these approaches. © Springer-Verlag Berlin Heidelberg 2007.