An Improved KNN Algorithm in Text Classification

Classification is make irregular tings regular according to different characteristics. Classification algorithm is a very important technology in data mining. The 21st is a century of information technology, people living in the world that has a variety of information. Classification techniques are used everywhere: we buy different products on different floors in the mall, different goods in different areas in supermarkets and so on. Also in industrial, aerospace, Internet and other areas, classification techniques can be seen. This article describes several existing classification algorithms. Their own characteristics were compared. As for traditional KNN algorithm's deficiencies when facing massive database has been improved. We proposed an improved KNN algorithm, and were compared in classification performance with the traditional KNN algorithm.