Research Of Text Clustering And Classification Method Based On Genetic Annealing Algorighms

The technology of Text classification and clustering is a very significant research topic in subjects of natural processing. We can promote the speed of query and retrieval for users by using Text classification and cluster, bringing much more convenient for us. So many academicians do much work in this field and make many achievements. But Text classification and clustering is after all a very complex and profound subject. A lot of problems still could be not solved currently. Based on this situation, the paper proposes some effective strategies to improve traditional algorithm of Text classification and clustering.The main research works in this paper can be interpreted as follows.1. As k-means Clustering Algorithm is excessively depend on the choice of the initial cluster centers, it leads to be involved in locally optimal solution and does not have comprehensive ability to search. The paper proposes a method based on Genetic algorithms and simulated annealing called Genetic Simulated Annealing algorithms. The new method has better Global convergence compared with Genetic algorithms. It also overcomes the aspect of premature convergence problem. So we can apply this algorithm to the k-means Clustering Method so as to overcome defect of k-means Clustering Method and optimize the result of cluster. Experiments indicated that this algorithm is efficient and accurate.2. As traditional KNN classification Algorithm has many defects, because of the low accuracy and speed of classification. The paper proposes a KNN classification method based on Genetic Simulated Annealing algorithms that we design. Experiments indicated that this algorithm is efficient and accurate.3. To solve the low accuracy of classification by KNN algorithm, this paper proposes an improving KNN algorithm based Genetic Simulated Annealing Clustering. This method can promote the accuracy of classification largely by combining the classification and clustering. Experiments indicated that this algorithm is efficient.4. Finally this paper design and develop a text classification system with k-means based Genetic Simulated Annealing algorithm and K-NN based Genetic Simulated Annealing algorithms. From the result of system’s running, these two improved algorithms are very effective.