Article Details

Abstract

To address the issues of strong subjectivity, low efficiency, and inconsistent standard execution prevalent in traditional organizational performance evaluation, an intelligent performance evaluation framework based on multi-model collaboration and repeatable verification is proposed. The framework formally defines the evaluation task as a function of rules, objects, and evaluator parameters. It designs a multi-model, multi-rule, and multi-round fusion strategy, effectively alleviating the inherent uncertainty and hallucination risks of Large Language Model (LLM) outputs through multiple sampling and statistical aggregation. Experiments conducted in three real-world scenarios of increasing difficulty including employee performance assessment, department party building evaluation, and project R&D evaluation, demonstrate that the framework not only provides stable quantitative scores but also synchronously generates structured qualitative feedback. This significantly improves the objectivity and efficiency of evaluation results, providing an effective technical solution for the intelligent and digital transformation of organizational management.

Keywords

Large Language Model; intelligent evaluation; performance assessment; multi-model collaboration; evaluation framework

References

1. [1] SUN J. Research on the Application of Large Language Models in Human Resource Management Practices [J].International Journal of Emerging Technologies and Advanced Applications，2024，1（8）：1-8.
2. [2] YERKEBULAN A，MANSUROVA M，ABDILDAYEVA A.Research on the Application of Large Language Models (LLM) for Improving Recruitment Efficiency and Accuracy [C]//Intelligent Information and Database Systems.Kitakyushu：Springer，2025：331-344.
3. [3] Wang Linchao Research on the Application of Intelligent Cognitive Large Models in Human Resource Management [J]. Enterprise Reform and Management, 2024 (20): 65-67
4. [4] Li Rong, Wang Qian, Li Yueqin Hospital Employee Performance Appraisal Model Based on Analytic Hierarchy Process and BP Neural Network [J]. Modern Scientific Instruments, 2023, 40 (6): 22-27
5. [5] Yang Cankui Application of Artificial Intelligence in Auxiliary Evaluation of Electric Power Materials [J]. Integrated Circuit Applications, 2023, 40 (4): 394-397
6. [6] Xiang Kun, Li Xiangfeng, Wang Renjie, etc Jingneng Procurement: Building the Best Practice of Intelligent Assisted Bid Evaluation Based on AI Large Model [J]. China Tendering, 2025 (S1): 103-113
7. [7] Zhu Fengyun, Zhang Guifeng, Yu Kang Research on Intelligent Assessment System for Vocational Skills Appraisal of Engineering Surveyors [J]. Modern Surveying and Mapping, 2018, 41 (3): 50-52
8. [8] Liao Guolong, Xiang Bazuoma, Li Hao, etc Research and Design of Medical Online Examination System Based on Agile Data Management Methodology: A Case Study of Beijing University of Traditional Chinese Medicine [J]. Traditional Chinese Medicine Education, 2024, 43 (3): 63-69
9. [9] Pan Ying Construction and Practice of Intelligent Assessment Workflow for Course Assignments Driven by Coze Platform [J]. Computer and Telecommunications, 2025 (8): 67-72
10. [10] Wang Qing, Wang Zhigang Research and Practice on Precision Teaching Mode of Higher Vocational Mathematics Course Based on AI Large Model [J]. Transportation Technology and Management, 2025, 6 (18): 186-188
11. [11] Zhu Shiling, Chen Lejun, Yuan Ziqian Design and Implementation of an Intelligent Examination System Integrating Multimodal Knowledge Analysis and Large Model Generation [J]. Modern Information Technology, 2025, 9 (20): 79-86, 91
12. [12] Lu Yu, Wu Yinghui Experimental Comparison of Seven Major Models in International Chinese Education Knowledge Assessment [J]. Journal of Hainan Normal University (Social Sciences Edition), 2025, 38 (3): 77-87
13. [13] Wu Haizhou, Wang Yuanyuan, Liu Jianing, etc Proposition Assistance System Based on Large Language Model [J]. Internet of Things Technology, 2025, 15 (20): 121-124128
14. [14] Li Weiqing, Wang Weijun, Huang Yinghui, etc Research on Post hoc Explanatory Recommendation Model Based on Generative Artificial Intelligence [J]. Journal of Intelligence, 2025, 44 (9): 1114-1127
15. [15] Chen Jiaoying, Bian Heshan A Skill Operation Assessment System Based on AI Technology [J]. Journal of Beibu Gulf University, 2020, 35 (10): 80-84
16. [16] Xie Likai, Yang Hui, Guo Wenxiu Research on Building an Intelligent Agent Cluster for State owned Assets Supervision Based on Large Models [J]. State owned Enterprise Management, 2025 (12): 111-113

About the authors

Feng Dingcheng (1984-), male, Han ethnicity, from Yancheng, Jiangsu, is an engineer with a PhD in engineering. His research interests include machine learning, complex networks, and project management; Corresponding author: Ma Ningfang (1975-), female, Hui ethnicity, Beijing native, senior engineer, master's degree, research direction: research on telecommunications legal regulation policies, communication network operation, and safety production management technology.