Information Extraction and Sentence Ordering in Multi-Document Summarization using Preference Learning

Abstract

Multi-document summarizing is a process that automatically extracts information from many texts that are related to the same subject. For the purpose of information extraction, a technique that uses multi-document summarization which is based on phrase frequency is used. As a result of the phrases being picked from the documents depending on how important they are, the summary loses their coherence and the sequence in which the information is presented, which reduces the readability of the summary. A method of sentence ordering that is predicated on the chronological order of the phrases has been used in order to address this issue. According to the findings of this research, a multi-document summarizer that is based on a word frequency approach performs very well when it comes to the process of extracting relevant content units and increasing the readability of the summary via sentence sequencing.