期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2014
卷号：62
期号：1
出版社：Journal of Theoretical and Applied
摘要：Research in text summarization is predominantly targets on measure of the worth of sentences for a summary. The proposed work has associated the Deep learning algorithm with fuzzy logic to improve the efficiency of the generated summary. The proposed work has two phases, they are training phase and testing phases. The training phase utilized to extract the benefits of fuzzy logic and deep learning algorithm for the efficient summary generation. Similar to every training phase, the proposed training phases is also possessed with well known data and attributes. Latter to the training phase, the testing phases is implemented to check the efficiency of the proposed approach. The experimental evaluation of the proposed work provided the predictable results as, the average precision obtained is 0.37, the average recall is 0.86 and the average f-measure is obtained as 0.50%.
关键词：Multi-document Summary; Deep Learning; Fuzzy Logic; Natural Language Processing; Neural Network.
