Automatic summarization of textual document

Abstract

In today world, there is a huge amount of information is growing every day on the internet and from many other sources and there is lots of textual information in it. To find out the relevant information from this large amount of data, we need an automatic mechanism that will extract the useful data. Such automatic systems are automatic summarization systems. They categorized into extractive and abstractive summarization system. Extractive summarization systems select the important sentences directly from the large document and put into summary whereas abstractive methods understand semantic meaning of the document by linguistic method to interpret and examine the text. In the purposed method, a statistical approach is used where multiple criterions or features are discussed to calculate the score for every sentence and then SIR (Susceptible Infected Recovered) model is used to compute the dynamic weight for every feature. After dynamic weight computation, weighted TOPSIS (The Technique for Order of Preference by Similarity to Ideal Solution) is used for multi-criterion analysis and aggregation. This method is fully implemented and integrated for automated textual document summarization system.