In order to get effective information timely and accurately in masses of text, text classification techniques get extensive attention from many aspects. A lot of algorithms were proposed for text classification which made it easy to classify texts, such as Naive Bayes, Rocchio, Decision Tree, Artificial Neural Networks, VSM, kNN and so on. In this paper, we mainly discussed the latest improved algorithm of kNN including Rocchio-kNN, TW-kNN, RS-kNN and kNN based on K-Medoids. Each of the representative algorithms is discussed in detail. These algorithms based on kNN have reduced the computational complexity as well as increased the execution efficiency compared with the traditional kNN algorithm.
展开▼
