Analysis of Research Trends and a Unified Framework for Comparative Text Summarization

The rapid expansion of unstructured text has intensified the need for effective and deployable text summarization systems. While extractive, abstractive, and large language model (LLM)–based approaches have advanced significantly, existing studies often evaluate these paradigms in isolation and lack standardized, reproducible frameworks. This study addresses this gap by combining a bibliometric analysis of LLM-related natural language processing (NLP) research with the development of a unified comparative summarization framework. A bibliometric analysis of 1,014 Scopus-indexed publications reveals key trends in LLM-based NLP research, including the dominance of conference-driven dissemination, concentration of research output among a small number of authors and institutions, and a strong geographic emphasis on Asia and North America. These findings highlight the rapid evolution of the field and the scarcity of deployable, standardized evaluation systems.