SUMMARIES: STRUCTURE, TOPIC, MAIN POINTS
https://doi.org/10.5281/zenodo.20378101
Kalit so‘zlar
extractive summarization, sentence scoring, intermediate representation, single-document summarization, multi-document summarization, text-to-text generation, abstractive summarization, information need.Annotasiya
This thesis examines extractive summarization systems and their operational principles. The topic focuses on how these systems generate concise, fluent summaries by identifying and concatenating salient sentences from single or multiple documents. The work is structured into three parts: core function and content selection, justification of extractive over abstractive methods, and three independent operational tasks. The main points include that extractive approaches are adaptable to users' information needs, function effectively for both single- and multi-document inputs, and involve three tasks: creating an intermediate representation, scoring sentences based on that representation, and selecting a summary consisting of several sentences. The introduction highlights critical design choices and explains how analyzing operational stages reveals advantages of certain techniques over others.
Foydalanilgan adabiyotlar ro‘yhati
Siddiqui, M. K., Ahmad, A., Pal, O., & Ahmad, T. (2021). CoRank: A clustering cum graph ranking approach for extractive summarization.
da Silva, V. C. (2022). Extractive text summarization using generalized additive models with interactions for sentence selection.
Yoon, S., Chan, H. P., & Han, J. (2023). PDSum: Prototype-driven continuous summarization of evolving multi-document sets stream. In ACM Web Conference 2023 (WWW '23) (pp. 1650–1661). Association for Computing Machinery.
Zhang, X., Wei, Q., Song, Q., & Zhang, P. (2024). TOMDS (Topic-Oriented Multi-Document Summarization): Enabling personalized customization of multi-document summaries. Applied Sciences, 14(5), 1880.
Jin, H., et al. (2024). A comprehensive survey on process-oriented automatic text summarization with exploration of LLM-based methods.
IEEE Access. (2025). Current trends and advances in extractive text summarization: A comprehensive review. IEEE Access, 13, 28150–28166.
