Integrating LDA with Clustering Technique for Relevance Feature Selection

Selecting features from documents that describe user information needs is challenging due to the nature of text, where redundancy, synonymy, polysemy, noise and high dimensionality are common problems. The assumption that clustered documents describe only one topic can be too simple knowing that most long documents discuss multiple topics. LDA-based models show significant improvement over the cluster-based in information retrieval (IR). However, the integration of both techniques for feature selection (FS) is still limited. In this paper, we propose an innovative and effective cluster- and LDA-based model for relevance FS. The model also integrates a new extended random set theory to generalise the LDA local weights for document terms. It can assign a more discriminative weight to terms based on their appearance in LDA topics and the clustered documents. The experimental results, based on the RCV1 dataset and TREC topics for information filtering (IF), show that our model significantly outperforms eight state-of-the-art baseline models in five standard performance measures.