Fuzzy genetic semantic based text summarization

Abstract

Automatic text summarization is a data reduction process to exclude unnecessary details and present important information in a shorter version. One way to summarize document is by extracting important sentences in the document. To select suitable sentences, a numerical rank is assigned to each sentence based on a sentence scoring approach. Highly ranked sentences are used for the summary. This paper proposed an automatic text summarization approach based on sentence extraction using fuzzy logic, genetic algorithm, semantic role labeling and their combinations to generate high quality summaries. This study explored the benefits of the genetic algorithm in the optimization problem in for feature selection during the training phase and adjusts feature weights during the testing phase. Fuzzy IF-THEN rules were used to balance the weights between important and unimportant features. Conventional extraction methods cannot capture semantic relations between concepts in a text. Therefore, this research investigates the use of the semantic role labeling to capture the semantic contents in sentences and incorporate it into the summarization method. This paper is evaluated in terms of performance using ROUGE toolkit. Experimental results showed that the summaries produced by the proposed approaches are better than other approaches produced by Microsoft Word 2007, Copernic Summarizer, and MANYASPECTS summarizers.