Research Of Text Classification Algorithm Based On The Quantum Immune Algorithm

Text classification is a method of the text mining which includes machine learning,statistical analysis, pattern recognition and information retrieval technologies and so on. Theclassification process generally includes text representation, construction and training ofclassifiers and evaluation of the classification results, in which the text representation can alsobe classified as text preprocessing, feature selection and weight calculation. In textclassification, we will get a large number of automatic classified texts. According to thedivided different classes, we can facilitate the use and management of the text. In addition,text classification as part of data mining and natural language processing, to deepen the study ontext classification, they also have extraordinary significance. So, for text classification researchand discussion is a highly practical value and theoretical significance.Text categorization is a complex problem of the natural language processing fieldinvolving many disciplines and professional knowledge. Many scholars have conductedin-depth research, but still text categorization exist many problems to be solved. This thesis isdirected by text object based on vector space model, which exists high dimensional and sparsedefects and thereby increase in the text classification time complexity and reduce in theperformance of text classification.Then, the paper is to develop a kind of text classificationalgorithm which can make full use of its high dimensional and sparse features. Therefore, thispaper will be applied to quantum immune algorithm in text classification, which can not onlydecreases the sparsely problem of text classification, and make better use of text vector highdimension feature.This paper first introduces the quantum computing knowledge. Secondly, discuss someknowledge on the immune algorithm. Then, carry on an in-depth study on the quantumimmune algorithm, and to make the appropriate adjustments and verify that it is better thanthe immune algorithm and quantum genetic algorithm. Finally, the quantum immunealgorithm specifically applied to text categorization, and proved by experiment, the quantumimmune algorithm based on text classification algorithm in some aspects of the classificationperformance is superior to other common text classification algorithm.