{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,10,10]],"date-time":"2025-10-10T01:37:06Z","timestamp":1760060226204,"version":"build-2065373602"},"reference-count":34,"publisher":"MDPI AG","issue":"8","license":[{"start":{"date-parts":[[2025,8,8]],"date-time":"2025-08-08T00:00:00Z","timestamp":1754611200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Symmetry"],"abstract":"<jats:p>Recent advancements in large language models (LLMs) have unlocked the potential for generating high-quality game commentary. However, producing insightful and engaging commentary for complex games with incomplete information remains a significant challenge. In this paper, we introduce a novel commentary method that combines reinforcement learning (RL) and LLMs, tailored specifically for the Chinese card game Guandan. Our system leverages RL to generate intricate card-playing scenarios and employs LLMs to generate corresponding commentary text, effectively emulating the strategic analysis and narrative prowess of professional commentators. The framework comprises a state commentary guide, a Theory of Mind (ToM)-based strategy analyzer, and a style retrieval module, which seamlessly collaborate to deliver detailed and context-relevant game commentary in the Chinese language environment. We empower LLMs with ToM capabilities and refine both retrieval and information filtering mechanisms. This facilitates the generation of personalized commentary content. Our experimental results demonstrate a significant improvement in the system\u2019s effectiveness in generating accurate, coherent, and engaging commentary when applied to open-source LLMs, surpassing GPT-4 across multiple evaluation metrics.<\/jats:p>","DOI":"10.3390\/sym17081274","type":"journal-article","created":{"date-parts":[[2025,8,8]],"date-time":"2025-08-08T09:56:42Z","timestamp":1754647002000},"page":"1274","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Enhancing Commentary Strategies for Guandan: A Study of LLMs in Game Commentary Generation"],"prefix":"10.3390","volume":"17","author":[{"given":"Jiayi","family":"Su","sequence":"first","affiliation":[{"name":"School of Computing and Data Science, Xiamen University Malaysia, Sepang 43900, Malaysia"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Meiling","family":"Tao","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Guangdong University of Technology, Guangzhou 510006, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Xuechen","family":"Liang","sequence":"additional","affiliation":[{"name":"Department of Computer Science, East China Jiaotong University, Nanchang 330013, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yangfan","family":"He","sequence":"additional","affiliation":[{"name":"Department of Computer Science & Engineering, University of Minnesota, Twin Cities, Minneapolis, MN 55414, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yiling","family":"Tao","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Sun Yat-sen University, Zhuhai 519082, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0000-0551-9678","authenticated-orcid":false,"given":"Miao","family":"Zhang","sequence":"additional","affiliation":[{"name":"Shenzhen International Graduate School, Tsinghua University, Shenzhen 518055, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2025,8,8]]},"reference":[{"key":"ref_1","unstructured":"Liao, J.W., and Chang, J.S. (1990, January 21\u201323). Computer Generation of Chinese Commentary on Othello Games. Proceedings of the Rocling III Computational Linguistics Conference III, Taipei, Taiwan."},{"key":"ref_2","unstructured":"Sadikov, A., Mo\u017eina, M., Guid, M., Krivec, J., and Bratko, I. (2006, January 29\u201331). Automated chess tutor. Proceedings of the Computers and Games: 5th International Conference, CG 2006, Turin, Italy."},{"key":"ref_3","doi-asserted-by":"crossref","unstructured":"Kameko, H., Mori, S., and Tsuruoka, Y. (September, January 31). Learning a game commentary generator with grounded move expressions. Proceedings of the 2015 IEEE Conference on Computational Intelligence and Games (CIG), Tainan, Taiwan.","DOI":"10.1109\/CIG.2015.7317930"},{"key":"ref_4","unstructured":"Guo, J., Yang, B., Yoo, P., Lin, B.Y., Iwasawa, Y., and Matsuo, Y. (2023). Suspicion-Agent: Playing Imperfect Information Games with Theory of Mind Aware GPT-4. arXiv."},{"key":"ref_5","unstructured":"Kim, B.J., and Choi, Y.S. (April, January 30). Automatic baseball commentary generation using deep learning. Proceedings of the the 35th Annual ACM Symposium on Applied Computing, Brno, Czech Republic."},{"key":"ref_6","unstructured":"Taniguchi, Y., Feng, Y., Takamura, H., and Okumura, M. (February, January 27). Generating live soccer-match commentary from play data. Proceedings of the AAAI Conference on Artificial Intelligence, Honolulu, HI, USA."},{"key":"ref_7","doi-asserted-by":"crossref","unstructured":"Zhao, Y., Lu, Y., Zhao, J., Zhou, W., and Li, H. (2023). DanZero+: Dominating the GuanDan Game through Reinforcement Learning. arXiv.","DOI":"10.1109\/TG.2024.3422396"},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Kosinski, M. (2023). Evaluating Large Language Models in Theory of Mind Tasks. arXiv.","DOI":"10.1073\/pnas.2405460121"},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"510","DOI":"10.1162\/tacl_a_00381","article-title":"Data-to-text generation with macro planning","volume":"9","author":"Puduppully","year":"2021","journal-title":"Trans. Assoc. Comput. Linguist."},{"key":"ref_10","doi-asserted-by":"crossref","unstructured":"Gardent, C., Shimorina, A., Narayan, S., and Perez-Beltrachini, L. (August, January 30). Creating training corpora for nlg micro-planning. Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, ACL 2017. Association for Computational Linguistics (ACL), Vancouver, BC, Canada.","DOI":"10.18653\/v1\/P17-1017"},{"key":"ref_11","unstructured":"Wang, Z., and Yoshinaga, N. (2022). Esports Data-to-commentary Generation on Large-scale Data-to-text Dataset. arXiv."},{"key":"ref_12","unstructured":"Wei, J., Bosma, M., Zhao, V.Y., Guu, K., Yu, A.W., Lester, B., Du, N., Dai, A.M., and Le, Q.V. (2021). Finetuned language models are zero-shot learners. arXiv."},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Ishigaki, T., Topi\u0107, G., Hamazono, Y., Noji, H., Kobayashi, I., Miyao, Y., and Takamura, H. (2021, January 20\u201324). Generating racing game commentary from vision, language, and structured data. Proceedings of the the 14th International Conference on Natural Language Generation, Scotland, UK.","DOI":"10.18653\/v1\/2021.inlg-1.11"},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Jhamtani, H., Gangal, V., Hovy, E., Neubig, G., and Berg-Kirkpatrick, T. (2018, January 15\u201320). Learning to generate move-by-move commentary for chess games from large-scale social forum data. Proceedings of the the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Melbourne, Australia.","DOI":"10.18653\/v1\/P18-1154"},{"key":"ref_15","doi-asserted-by":"crossref","unstructured":"Nimpattanavong, C., Taveekitworachai, P., Khan, I., Nguyen, T.V., Thawonmas, R., Choensawat, W., and Sookhanaphibarn, K. (2023, January 6\u20139). Am I Fighting Well? Fighting Game Commentary Generation With ChatGPT. Proceedings of the the 13th International Conference on Advances in Information Technology, Bangkok, Thailand.","DOI":"10.1145\/3628454.3629551"},{"key":"ref_16","unstructured":"Shahul, E., James, J., Anke, L.E., and Schockaert, S. (2023). RAGAS: Automated Evaluation of Retrieval Augmented Generation. arXiv."},{"key":"ref_17","first-page":"1","article-title":"Improving the Domain Adaptation of Retrieval Augmented Generation (RAG) Models for Open Domain Question Answering","volume":"11","author":"Siriwardhana","year":"2022","journal-title":"Trans. Assoc. Comput. Linguist."},{"key":"ref_18","doi-asserted-by":"crossref","unstructured":"Chen, W., Hu, H., Chen, X., Verga, P., and Cohen, W.W. (2022). MuRAG: Multimodal Retrieval-Augmented Generator for Open Question Answering over Images and Text. arXiv.","DOI":"10.18653\/v1\/2022.emnlp-main.375"},{"key":"ref_19","unstructured":"Melz, E. (2023). Enhancing LLM Intelligence with ARM-RAG: Auxiliary Rationale Memory for Retrieval Augmented Generation. arXiv."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"251","DOI":"10.1016\/0893-6080(91)90009-T","article-title":"Approximation capabilities of multilayer feedforward networks","volume":"4","author":"Hornik","year":"1991","journal-title":"Neural Networks"},{"key":"ref_21","doi-asserted-by":"crossref","unstructured":"Horn, R.A., and Johnson, C.R. (1985). Matrix Analysis, Cambridge University Press.","DOI":"10.1017\/CBO9780511810817"},{"key":"ref_22","unstructured":"Rudin, W. (1976). Principles of Mathematical Analysis, McGraw-Hill. [3rd ed.]."},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"R644","DOI":"10.1016\/j.cub.2005.08.041","article-title":"Theory of mind","volume":"15","author":"Frith","year":"2005","journal-title":"Curr. Biol."},{"key":"ref_24","unstructured":"Achiam, J., Adler, S., Agarwal, S., Ahmad, L., Akkaya, I., Aleman, F.L., Almeida, D., Altenschmidt, J., Altman, S., and Anadkat, S. (2023). Gpt-4 technical report. arXiv."},{"key":"ref_25","unstructured":"Bai, J., Bai, S., Chu, Y., Cui, Z., Dang, K., Deng, X., Fan, Y., Ge, W., Han, Y., and Huang, F. (2023). Qwen technical report. arXiv."},{"key":"ref_26","unstructured":"Young, A., Chen, B., Li, C., Huang, C., Zhang, G., Zhang, G., Li, H., Zhu, J., Chen, J., and Chang, J. (2024). Yi: Open Foundation Models by 01.AI. arXiv."},{"key":"ref_27","unstructured":"Zeng, A., Liu, X., Du, Z., Wang, Z., Lai, H., Ding, M., Yang, Z., Xu, Y., Zheng, W., and Xia, X. (2022). Glm-130b: An open bilingual pre-trained model. arXiv."},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"130","DOI":"10.1108\/eb046814","article-title":"An algorithm for suffix stripping","volume":"14","author":"Porter","year":"1980","journal-title":"Program"},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"45","DOI":"10.1016\/S0306-4573(02)00021-3","article-title":"An information-theoretic perspective of tf\u2013idf measures","volume":"39","author":"Aizawa","year":"2003","journal-title":"Inf. Process. Manag."},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"1093","DOI":"10.1016\/j.asej.2014.04.011","article-title":"Sentiment analysis algorithms and applications: A survey","volume":"5","author":"Medhat","year":"2014","journal-title":"Ain Shams Eng. J."},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"201","DOI":"10.1017\/S0305000900012885","article-title":"Type\/token ratios: What do they really tell us?","volume":"14","author":"Richards","year":"1987","journal-title":"J. Child Lang."},{"key":"ref_32","unstructured":"isnowfy (2025, March 28). SnowNLP. Available online: https:\/\/github.com\/isnowfy\/snownlp."},{"key":"ref_33","unstructured":"Renella, N., and Eger, M. (2023, January 8). Towards automated video game commentary using generative AI. Proceedings of the the AIIDE Workshop on Experimental AI in Games, Salt Lake City, UT, USA."},{"key":"ref_34","doi-asserted-by":"crossref","unstructured":"Lin, C.Y., and Hovy, E. (2002, January 11\u201312). Manual and automatic evaluation of summaries. Proceedings of the the ACL-02 Workshop on Automatic Summarization. Association for Computational Linguistics, Philadelphia, PA, USA.","DOI":"10.3115\/1118162.1118168"}],"container-title":["Symmetry"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2073-8994\/17\/8\/1274\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,9]],"date-time":"2025-10-09T18:26:16Z","timestamp":1760034376000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2073-8994\/17\/8\/1274"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,8,8]]},"references-count":34,"journal-issue":{"issue":"8","published-online":{"date-parts":[[2025,8]]}},"alternative-id":["sym17081274"],"URL":"https:\/\/doi.org\/10.3390\/sym17081274","relation":{},"ISSN":["2073-8994"],"issn-type":[{"type":"electronic","value":"2073-8994"}],"subject":[],"published":{"date-parts":[[2025,8,8]]}}}