Fuzzy Information Retrieval from Mining Relational Database by Using Link Analysis Mining Methods
M. Sivanjaneyulu, A. Anuradha · 2012
Link Analysis algorithms have been powering various search engines for efficient web information retrievals. Instead of web, in this paper we propose to use Link Analysis as an extension of correspondence analysis in a relational database for its ability to effectively discover relationships. Initially, a reduced, much smaller, Markov chaining containing only the elements of interest is extracted and refined by stochastic complementation. This reduced chain is then analyzed by projecting jointly the elements (entity relations in relational database) of interest in a kernel version of the diffusion-map subspace along with spectral clustering to visualize the results. Also applying this technique for fuzzy information retrievals can improve overall performance in a relational database. Experiments show the usefulness of the technique for extracting relationships in relational databases.