A new Text Knowledge Element Mining model Based on Knowledge Spectrum Computing

Youkui Wen, School of Economy and Management Xidian Univerisity at Xi'an in China

This paper present a new text feature extraction model: Text Knowledge Elements Mining (TKEM) that based on the concepts of Knowledge spectrum transformation for information. We can decompose the text knowledge into independent knowledge elements, Which is consist of guidance information and it knowledge spectrum external entity. Discovery of knowledge element is realized by computing the relativity between guidance information and knowledge spectrum external entity. and experimental results show TKEM is a superior model.