Jinan, Shandong – The 24th China National Conference on Computational Linguistics (CCL 2025) has released its evaluation results. A team from the Open University of China (OUC) won the first place in the “Chinese Essay Rhetoric Recognition Evaluation” task.

The team comprised  LAI Yuxuan and NIU Zequn from the School of Science and Engineering, WANG Xiajing from the Experimental College, ZHANG Huan from the Department of Digitalisation, and ZHENG Chen from  the Engineering Research Centre. They competed against over 20 teams from institutions such as Beijing Institute of Technology, Xiamen University and Beijing Normal University.

Hosted by the Chinese Information Processing Society of China, CCL is the country’s flagship natural language processing conference. This year’s “Chinese Essay Rhetoric Recognition Evaluation” task featured three sub-tracks: rhetorical form classification, rhetorical content classification, and rhetorical element extraction. These were designed to evaluate the performance of large language models in identifying rhetorical devices in student writing within real educational scenarios.

Amid strong competition from many top universities, the OUC team demonstrated solid technical expertise. By applying techniques such as large-model fine-tuning, context learning, and model ensemble, they significantly improved the accuracy and generalization ability of the model across different rhetorical tasks. The team achieved highest score in all three sub-tracks, ranking at the top overall.

This achievement validates the effectiveness of the OUC team’s technical approach. Going forward, the team will continue to explore educational digitalisation technologies, develop smarter tools to assist teaching and learning, and advance AI-driven innovation and reform in education.

By ZHENG Chen, OUC 

Source: OUC Official WeChat Account