Memory Principle Based Domain Word Relationship Network Construction Model

Automatic domain knowledge extraction is one of the key problems in text processing. To construct an effective domain word relationship network is important to text mining, text clustering and text information retrieving. A segmentation algorithm based on famous Apriori algorithm is issued and then a new construction model of domain word relationship network is proposed to organize domain words occurred in the same context with human memory principle. Some experiment results show that this model can extract domain words and construct their relationships from a great deal of domain documents effectively