A New Term Ranking Method based on Relation Extraction and Graph Model for Text Classification

Huynh, D., Tran, D. and Ma, W.

    Term frequency and document frequency are currently used to measure term significance in text classification. However, these measures cannot provide sufficient information to differentiate important terms. Thus, in this research, a new term ranking (weighting) approach for text classification will be proposed. The approach firstly is based on relations among terms to estimates the important levels of terms in a document. Secondly, the proposed approach provides a considerable representation for the text documents. The results from experiment show that with the same data in Wikipedia corpus the term weighting approach provides higher accuracy in comparison to the popular approaches based on term frequency.
Cite as: Huynh, D., Tran, D. and Ma, W. (2011). A New Term Ranking Method based on Relation Extraction and Graph Model for Text Classification. In Proc. Australasian Computer Science Conference (ACSC 2011) Perth, Australia. CRPIT, 113. Mark Reynolds Eds., ACS. 145-152
pdf (from crpit.com) pdf (local if available) BibTeX EndNote GS