Sentence Extraction Based Single Document Summarization

The need for text summarization is crucial as we enter the era of information overload. In this paper we present an automatic summarization system, which generates a summary for a given input document. Our system is based on identification and extraction of important sentences in the input document. We listed a set of features that we collect as part of summary generation process. These features were stored using vector representation model. We defined a ranking function which ranks each sentence as a linear combination of the sentence features. We also discussed about discourse coherence in summaries and techniques to achieve coherent and readable summaries. The experiments showed that the summary generated is coherent the selected features are really helpful in extracting the important information in the document.