Abstract
In recent years, there has been a tremendous growth of online text information related to digital libraries, medical diagnostic systems, remote education, news sources and electronic commerce. There is a great need to search and organize huge amount or information in text documents. This paper focuses on word tendencies in documents and suggests an automatic extraction method for specific subject. Field judgment is conducted by using field association words, and similarity, among word tendencies and other word tendencies is computed with information. Then word tendencies which have the same subject are grouped as one group and the important word tendencies are chosen from that group. Finally, a system suggests word tendencies from specific subjects, and fields are implemented