数字图书馆：2025–2026 年度综述

An annual survey of the cs.DL arXiv category, covering research published from September 2025 – August 2026. Built on a close-reading set of 108 papers and organised into 40 chapters, it maps the year by the field's research sub-directions.

Book framework (share of the year's output):

• Semantic Metadata and Knowledge Graph Infrastructure (24.0%) — Focuses on enhancing digital library retrieval and organization through semantic metadata, knowledge graphs, and multimodal datasets for historical and scientific content.
• Scientometric Analysis of Research Output and Impact (14.2%) — Employs bibliometric and scientometric methods to evaluate research productivity, citation patterns, and institutional performance across various scientific disciplines.
• Library and Information Science Methodological Evolution (12.5%) — Analyzes the historical development and current trends of research methods and topic convergence within the Library and Information Science discipline.
• Large Language Models in Scientific Literature Processing (12.2%) — Investigates the application of Large Language Models for extracting entities, mapping topics, and supporting workflows in scientific literature analysis.
• AI-Assisted Scholarly Workflows and Research Diversity (10.1%) — Examines the impact of generative AI on scholarly writing, review processes, and research diversity, including issues of bias and visibility in AI-assisted science.
• Citation Network Integrity and Hallucination Verification (9.0%) — Addresses challenges in citation network analysis, specifically focusing on detecting hallucinated citations and verifying the structural integrity of scholarly references.
• Scientific Novelty, Impact, and Paradigm Dynamics (8.5%) — Explores metrics and semantic frameworks to identify paradigm-shifting discoveries, assess scientific novelty, and understand the dynamics of knowledge creation.
• Peer Review Automation and Scholarly Evaluation (6.7%) — Studies the automation of peer review processes, reviewer behavior analysis, and the use of AI tools to improve the efficiency and fairness of scholarly evaluation.
• Research Software Metadata and Reproducibility Infrastructure (2.9%) — Develops tools and schemas for managing research software metadata, ensuring reproducibility, and linking software artifacts to scholarly outputs.

Each sub-direction is covered across three chapters — how it framed its questions, the strength of its key findings, and its open debates and frontiers — so the reader gets a traceable picture of what this part of the field actually produced over the year.