A Constraints Driven PSO Based Approach for Text Summarization

In the present scenario we are living in a digital media and virtual world. To conveniently communicate in digital world electronic data have to gradually increase. So it is a serious challenge to manage the huge digital and electronic resources efficiently and accurately. One of the important solutions of the above problem is text summarization i.e. an application of text mining. Representing the gist of a text document is called summary. A rich summary always covers the maximum coverage, high level of diversity and with user defined size. This paper proposes an approach for summarizing the text documents by extractive way using Particle Swarm Optimization (PSO) that is known as population based stochastic optimization technique and it has many similarities with evolutionary computation techniques such as Genetic Algorithms (GA). The huge volume and dimensions of terms have managed by the concepts of term document matrix followed by K-Means clustering with PSO for acquiring optimal number of concepts clusters. Then apply constraint-driven concept for selecting the best one. These key concepts were used to identify the significant gist in documents for text summarization.