Organizing Web Documents Resulting from an Information Retrieval System Using Formal Concept Analysis

To discover information needs among information resulting from an information retrieval (IR) system by a user, it is needed to be managed them in some effective ways. Document clustering is a common and useful technique for Web information retrieval. In this paper, we use formal concept analysis (FCA) method for reorganizing documents resulting from an IR system according to their formal concepts. We use tf.idf (term frequency x inverse document frequency) term weighting scheme in selecting terms from the documents to construct the formal context of documents and terms which will then be used to extract formal concepts among resulting documents. We use text mining technique, association rule mining, on the frequent termsets of the given domain to analyze relationships of terms existing in the resulted documents. Finally, the concept lattice of the documents is built on these associated terms to obtain formal concept based clustered resulted documents by discovering ordering relations among frequent termsets using lattice theory