Text Classification Methods Based on SVD and FCM

摘要

In order to find key and useful messages among massive online resources, this paper propose a method to classify documents about soybean metabolism based on Singular Value Decomposition (SVD) and Fuzzy c-Means(FCM). Singular Value Decomposition (SVD) is an important way of matrix decomposition, which can represent a complex matrix by dividing it into smaller and simpler submatrices that describe important properties of matrices. After the dimension reduction, the Fuzzy c-Means (FCM) is used for clustering, which makes the objects divided into the same cluster have the highest similarity, while the object between different clusters have the lowest similarity. Besides, term frequency (TF) and entropy weight method (EWM) can also be used to construct matrix.