A Hybrid Method for Query based Automatic Summarization System

Automatic text summarization is one of the research goals of Natural Language Processing which relieves humans from studying each and every line in a text document to understand the underlying concepts in it. Automatic text summarization is aimed to create a brief outline of a given text covering the important points in the text. Automatic text summarization can be generic or query specific. This paper is focused on Query specific text summarization where a summary of the given text is constructed based on the given query. Query specific text summarization is based on the calculation of the relationship between sentences in the text document and the query given. Several statistical techniques and linguistic techniques have been developed to find the relationship between the given query and the sentences in the document. These methods when used alone could not give desired accuracy in the results. In this paper a sentence scoring method is defined based on existing sentence scoring methods. It attempts to combine the individual results of these methods to give a better assessment of the relationship between the sentences.