Performance Analysis of Frequency and Graph Theoretic Based Text Summarization

Abstract

Nowadays people are becoming busy with their office and family activities. They do not get enough time to study the newspaper completely. But they need to update themselves to the events of the current world. Hence, text summarization can help them to know the events happening all over the world very quickly and saves our valuable time. Automatic summarization is a way of bringing that gist from a large paragraph by coding in the computer. In the paper, we have used term frequency and graph theory summarizing techniques where the first uses word frequency while in the second, each sentence is assumed as a node and the resemblance between sentences is assigned to their corresponding edges. Here, we have used python as the programming language for coding. The analysis shows that the graph-theoretic approach provides more accurate result than frequency summarizer although it takes more time.