Spoken query processing for interactive information retrieval

It has long been recognised that interactivity improves the effectiveness of information retrieval systems. Speech is the most natural and interactive medium of communication and recent progress in speech recognition is making it possible to build systems that interact with the user via speech. However, given the typical length of queries submitted to information retrieval systems, it is easy to imagine that the effects of word recognition errors in spoken queries must be severely destructive on the system's effectiveness. The experimental work reported in this paper shows that the use of classical information retrieval techniques for spoken query processing is robust to considerably high levels of word recognition errors, in particular for long queries. Moreover, in the case of short queries, both standard relevance feedback and pseudo relevance feedback can be effectively employed to improve the effectiveness of spoken query processing.