Machine Learning Algorithms with Co-occurrence Based Term Association for Text Mining

In this paper, an effective method for computing term association from a text corpus is presented. Two machine learning algorithms are employed to evaluate the effectiveness of the proposed method for text mining. The co-occurrence based term association method is to overcome the problem of lack of relationship between words for keyword based text mining and improve the performance of text mining. The experiments are conducted on the standard Reuter-21578 data set and 20 news group data set. Different number of associated terms are compared in the experiments. Experimental results show that the proposed method can achieve better results on different machine learning algorithms when measured by F measure.