An algebraic approach for sentence based feature extraction applied for automatic text summarization
In the existing approaches of Automatic Text Summarization (ATS) which uses the algebraic reduction method Non-negative Matrix Factorization (NMF), the approximation process is more concerned towards convergence and feature selection from the original text into W matrix is of less consideration. Thi...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English English |
| Published: |
American Scientific Publishers
2014
|
| Subjects: | |
| Online Access: | http://irep.iium.edu.my/41836/ http://irep.iium.edu.my/41836/ http://irep.iium.edu.my/41836/ http://irep.iium.edu.my/41836/3/ASP_NAA_FINAL.pdf http://irep.iium.edu.my/41836/6/41836_An%20algebraic%20approach_SCOPUS.pdf |
| Summary: | In the existing approaches of Automatic Text Summarization (ATS) which uses the algebraic reduction method Non-negative Matrix Factorization (NMF), the approximation process is more concerned towards convergence and feature selection from the original text into W matrix is of less consideration. This obviously leads to linguistic noise which affects the quality of the summary generated the ATS process. Each sentence in a given text plays a vital role and identifying the most relevant and an important sentence from the given text is most prominent for ATS. This can be done when only the correct features of the sentences are extracted properly in W matrix. Hence this paper proposes an algebraic approach for sentence based feature extraction mainly addressing the issue that exists with the Non-negative Matrix Factorization (NMF) based methods when applied to ATS. Result clearly indicates that the newly proposed approach can identify and extract feature terms from documents effectively and weigh them more accurately than the existing method addressed. |
|---|