A Comparative Study of Computational and Linguistic Approaches to Summarization

Abstract

In the digital age, summarization is a vital tool for information management, condensing vast quantities of data into manageable insights. This paper examines two primary approaches to summarization—computational and linguistic—each with distinct methodologies and strengths. Computational techniques, including extractive and abstractive methods, prioritize efficiency and scalability, employing algorithms and advanced neural networks to generate summaries. Linguistic approaches, however, treat summarization as an interpretive process, focused on preserving intent, coherence, and the communicative goal of the text. This paper compares these two paradigms, exploring their respective advantages and challenges. Ultimately, we argue that integrating computational models with linguistic principles offers a more robust framework for generating human-centered, contextually aware summaries