Scholar4AI Updated 2026-10-26 14:32 UTC+09:00

Research published in October 2025

What the tracked finance, accounting, economics, management, operations research and AI journals published in October 2025, by publication date in OpenAlex. This page is a fixed monthly record: its address does not change. Counts can still grow a little as late papers are indexed, and citations are updated daily. Also as Markdown.

Share: X LinkedIn Bluesky Threads Facebook Reddit Hacker News Hatena LINE Email
Copy:

← September 2025 · All months · November 2025 →

Key findings#

  • In October 2025, the tracked journals published 3,737 papers (finance 621, accounting 49, economics 801, management 395, operations research 491, AI 1,380).
  • The most frequent keywords were “China” (133), “Chinese listed companies” (112) and “large language models” (83).
  • Compared with their share in the twelve months before, the keywords that grew most were “explainable AI” (42 papers, +17), “digital infrastructure” (20 papers, +14) and “AI ethics” (22 papers, +12).
  • The institutions with the most papers were Chinese Academy of Sciences (55), Tsinghua University (54) and Sun Yat-sen University (46).
  • China was involved in 48% of the papers.
  • The most-cited paper published that month is “Theory and theory development: Guidelines for establishing theoretical gaps, foundations, contributions, and implications” (Journal of Business Research), with 165 citations so far.

Papers by category#

Papers published in October 2025
CategoryPapers
Finance621
Accounting49
Economics801
Management395
Operations Research491
AI1,380

Top keywords#

Most frequent keywords, October 2025
#KeywordPapers
1China133
2Chinese listed companies112
3large language models83
4corporate governance83
5financial constraints71
6state-owned enterprises65
7COVID-19 pandemic62
8graph neural networks61
9digital transformation58
10regional heterogeneity50
11moderating effect45
12explainable AI42
13green innovation41
14artificial intelligence40
15contrastive learning40

Growing keywords#

Papers in October 2025 beyond the keyword's share in the twelve months before
KeywordPapersAt earlier shareExtra papers
explainable AI4225.0+17.0
digital infrastructure206.5+13.5
AI ethics229.9+12.1
diffusion models3118.9+12.1
generative AI2917.8+11.2
information transparency2110.1+10.9
mechanism design154.7+10.3
digital twin166.5+9.5
domain shift177.6+9.4
inflation2415.2+8.8

Top keywords by category#

The five most frequent keywords per category, October 2025
CategoryMost frequent keywords
FinanceChinese listed companies (77); corporate governance (55); China (53); state-owned enterprises (46); financial constraints (44)
Accountingcorporate governance (12); Chinese listed companies (4); financial constraints (3); tax avoidance (3); 10-K filings (2)
EconomicsChina (68); inflation (21); COVID-19 pandemic (20); monetary policy (19); income inequality (18)
Managementdigital transformation (18); Chinese listed companies (15); small and medium-sized enterprises (12); corporate social responsibility (10); dynamic capabilities (10)
Operations Researchmixed-integer linear programming (29); multi-objective optimization (14); distributionally robust optimization (13); COVID-19 pandemic (11); adaptive large neighborhood search (11)
AIlarge language models (66); graph neural networks (60); contrastive learning (40); class imbalance (34); explainable AI (33)

Institutions#

Institutions with the most papers, October 2025
#InstitutionPapers
1Chinese Academy of Sciences55
2Tsinghua University54
3Sun Yat-sen University46
4Huazhong University of Science and Technology45
5University of Electronic Science and Technology of China45
6Hong Kong Polytechnic University41
7Peking University41
8Zhejiang University40
9Harbin Institute of Technology38
10Tongji University38

Countries#

Countries, October 2025
#CountryPapers
1China1,779
2United States579
3United Kingdom296
4Australia175
5Germany175
6Italy165
7France139
8Hong Kong131
9India131
10Canada120

Journals#

Equations and models#

Datasets#

Most-mentioned datasets, October 2025 (titles, abstracts and keywords)
#DatasetPapers
1ImageNet7
2CIFAR-10 / CIFAR-1005
3MS COCO4
4MNIST / Fashion-MNIST2
5MuJoCo2
6YouTube / TikTok2
7ADE20K1
8Cora / CiteSeer / PubMed (Planetoid)1
9Face recognition (LFW, CelebA, VGGFace2)1
10Facebook1

Most-cited papers published in October 2025#

Most-cited papers published in October 2025 (citations counted by OpenAlex, updated daily)
#CitationsPer monthTitleJournalPublishedAuthors
116513.8Theory and theory development: Guidelines for establishing theoretical gaps, foundations, contributions, and implicationsJournal of Business Research2025-10-27Weng Marc Lim
211810.0Datasets for large language models: a comprehensive surveyArtificial Intelligence Review2025-10-31Yang Liu, Jiahuan Cao, Chongyu Liu et al.
31179.5Safeguarding large language models: a surveyArtificial Intelligence Review2025-10-17Yi Dong, Ronghui Mu, Yanghao Zhang et al.
4645.3Applications, classifications, and challenges: a comprehensive evaluation of recently developed metaheuristics for search and analysisArtificial Intelligence Review2025-10-23Muhammad Suhail Shaikh, Saurav Raj, Gengzhong Zheng et al.
5615.0A survey on retrieval-augmentation generation (RAG) models for healthcare applicationsNeural Computing and Applications2025-10-16Mohamed Ahmed Abo El-Enen, Sally Saad, Taymoor Mohamed Nazmy
6584.7CNNs, RNNs and Transformers in human action recognition: a survey and a hybrid modelArtificial Intelligence Review2025-10-17Khaled Alomar, Halil Ibrahim Aysel, Xiaohao Cai
7514.3Intelligent financial system: How AI is transforming financeJournal of Financial Stability2025-10-29Iñaki Aldasoro, Leonardo Gambacorta, Anton Korinek et al.
8494.0A Comprehensive Survey of Dynamic Graph Neural Networks: Models, Frameworks, Benchmarks, Experiments and ChallengesIEEE Transactions on Knowledge and Data Engineering2025-10-14ZhengZhao Feng, Rui Wang, TianXing Wang et al.
9483.8Sample-efficient backtrack temporal difference deep reinforcement learningKnowledge-Based Systems2025-10-10Qi Liu, Pengbin Chen, Ke Lin et al.
10483.9Resolving data bias improves generalization in binding affinity predictionNature Machine Intelligence2025-10-21David Graber, Peter Stockinger, Fabian Meyer et al.

← September 2025 · All months · November 2025 →

Cite this page

Scholar4AI. (2026). Research published in October 2025. Journal Research Rankings. Retrieved October 26, 2026, from https://scholar4ai.com/reports/2025-10

@misc{scholar4ai-reports-2025-10,
  author = {{Scholar4AI}},
  title = {Research published in October 2025},
  year = {2026},
  howpublished = {\url{https://scholar4ai.com/reports/2025-10}},
  urldate = {2026-10-26},
  note = {Journal Research Rankings; data from OpenAlex (CC0); licensed CC BY-NC 4.0}
}

Free to reuse for non-commercial purposes with attribution (CC BY-NC 4.0).