Fuzzy K-means document clustering analysis based on PSO algorithm
Lin Quan · Jisuanji gongcheng yu sheji · 2009
Document clustering has aroused broad attention with the rapid increase in the amount of the web text and the needs in practical application and it rapidly becomes a research subject in artificial intelligence.The keyword membership is defined as a ratio of the fre-quency of keyword shown in the document to the length of the document.N keywords which is shown most frequently in M pieces of documents is regarded as the features of document membership.And M document memberships compose a model sample set of document fuzzy clustering.Particle Swarm Optimization algorithm is embedded into the fuzzy K-mean clustering algorithm to optimize the total scattering degree of clusters to be the minimum and to obtain the optimization of documents clusters.The results of simulative experiments indicate that the application of PSO algorithm to the document clustering problems can obtain better clusters results.