Measuring the Quality of Text Summarization: A Survey of Evaluation Approaches

Abstract

Text summarization and text summarization measures both play significant roles in the field of natural language processing and information retrieval. Text summarization is a computational technique that condenses large volumes of text into concise summaries, extracting key information and main ideas. It enables users to efficiently retrieve relevant information, comprehend complex documents, and make informed decisions. On the other hand, text summarization measures are essential tools for evaluating the quality, effectiveness, and performance of text summarization systems. These measures encompass various aspects, including content coverage, linguistic quality, coherence, and informativeness of the generated summaries. By leveraging text summarization measures, researchers and practitioners can systematically evaluate, benchmark, and improve summarization methods. They facilitate the identification of best practices, the development of innovative techniques, and the fair comparison of different approaches. This paper presents a comprehensive comparative analysis of various text summarization measures. © 2023 IEEE.