A K-mixture connective-strength-based approach to automatic text summarisation
by Te-Min Chang, Wen-Feng Hsiao
International Journal of Intelligent Systems Technologies and Applications (IJISTA), Vol. 10, No. 2, 2011

Abstract: This research focuses on developing a hybrid automatic text summarisation approach, KCS, to enhance the quality of summaries. KCS employs the K-mixture probabilistic model to establish term weight distributions in a statistical sense. It further identifies the lexical relations between nouns and nouns, as well as nouns and verbs to derive the connective strength (CS) of nouns. Sentences are ranked and extracted according to the accumulated CS values they contain. We conduct two experiments to justify the proposed approach. The results show that the K-mixture model itself is more conducive to document classification than traditional TFIDF weighting scheme since the best macro F-measure increases from 0.63 to 0.67. It, however, is still no better than the more complex linguistic-based approach that takes noun's CS into consideration. Most importantly, our proposed approach, KCS, performs best among all approaches considered (with the best macro F-measure of 0.8). It implies that KCS can extract more representative sentences from the document and its feasibility in text summarisation applications is thus justified.

Online publication date: Fri, 11-Mar-2011

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Intelligent Systems Technologies and Applications (IJISTA):
Login with your Inderscience username and password:

    Username:        Password:         

Forgotten your password?


Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com