Combining Multiple Evidence from Different Relevant Feedback Networks

It has been known that using di’erent repre-sentations of a query retrieves different sets of documents. Recent work suggests that sig-nificant improvement in retrieval performance can be achieved by combining multiple repre-sentations of an information need. In this paper, we first investigate a fully automatic way of generating multiple query representa-tions for (I given information problem. We produce multiple query vectors by expanding an initial query vector with various relevance feedback methods. We then describe the effect of combining the multiple query vectors on retrieval effectiveness. Experimental results show that significant improvements can be obtained by the combination of multiple query vectors expanded with different relevance feed-back methods