Information retrieval has gained momentum due to increase in the various optimization algorithms. Information retrieval based on the user query needs challenging algorithms, since the features may vary between the user request and the documents in the database. This work introduces information retrieval system based on the clustering of the documents. The proposed approach performs the information retrieval through pre-processing, feature selection, and clustering. The proposed model uses the ACO algorithm for the feature selection and the k-means clustering algorithm for the clustering approach. When the query is provided by the user, the proposed model checks the similarity between the query and the each cluster and thus provides the documents in the cluster most similar to the query. The simulation uses two standard databases such as TREC, and the 20 newsgroup for the processing. From the simulation results, it is evident that the proposed ACO with the k-means algorithm has the improved accuracy 0.6105 and 0.4444 for the TREC and the 20 newsgroup database respectively. Similarly, the proposed model has reduced fallout value of 0.1505 and 0.165 for the TREC and the 20 newsgroup database respectively.
