Investigations on document summarization using cognitive models

Abstract

Information content available for use in various real world scenarios and applications has greatly increased recently and continues to grow. Specifically, the huge growth of textual information available in the Internet and intranets has made finding relevant text information on time as a major technical challenge. Document summarization is a reductive transformation of input texts into a summary text and can help in comprehending the abundant textual information. A good document summarization system should preserve the salient information content in the texts. Document summarization can be classified into various types based on the context factors of the summarization problem. Many document summarization techniques exist. However, the effectiveness of these systems, in terms of quality of the summaries created, is found lagging when compared to the quality of the summaries created by human beings. Very few document summarization techniques make use of human information processing behaviour and even such systems have severe limitations and strict assumptions. Human cognition and intelligent information processing behaviour are described by cognitive psychology models. Computational cognitive models transform the cognitive psychology models to adapt them for real world applications. Computational cognitive models for document summarization built from cognitive psychology models could address the gap between the effectiveness of system generated summaries and manually created summaries. newline