Combinational Semantic Approach for Efficient Data Extraction from Multiple Documents

Abstract

As the availability of data increase in the form of web pages, there arises a challenge of effective processing of data in a timely manner. Millions of documents were added day by day in World Wide Web, and for a manual processing it will take thousands of years and hence the term automatic data extraction comes to the context. On these gigantic volumes, textual data holds a major portion and effective text processing algorithms are needed for the processing and acquisition of data. Even though a number of techniques are available for sentence extraction no one can perform well as that of a human expert. But semantic approaches can perform better than other existing approaches, since they are considering the meanings rather that its form. Here the similarity calculations are more accurate than other approaches and the level of accuracy depends on the semantic tool they have used and the efficiency of the logic used for extracting the meaning, calculating similarity etc. In this proposal we are presenting a combinational approach of statistical and semantic procedures for sentence extraction which mainly differs from the previous approaches in the use of the semantic tool ThemeSets.