HyperAIHyperAI

Command Palette

Search for a command to run...

Three Research Achievements from the Department of Computer Science Win the 2024 Qian Weichang Chinese Information Processing Science and Technology Award - Tsinghua University

### Abstract: Three Research Projects from Tsinghua University's Computer Science Department Win 2024 Qian Weichang Chinese Information Processing Science and Technology Award On December 17, 2024, the 2024 Chinese Information Processing Society (CIPS) Annual Conference and the Third National Large Model Intelligent Generation Conference (CIPS-LMG 2024) took place in Jiaxing, Zhejiang. During the conference, the 2024 Qian Weichang Chinese Information Processing Science and Technology Award was presented. Three projects from Tsinghua University's Computer Science Department were honored, reflecting significant advancements in content recommendation, legal intelligence, and natural language generation. **1. Responsible Content Recommendation: Natural Science First Prize** Professor Min Zhang's team, in collaboration with Kuaishou Company, received the Natural Science First Prize for their project titled "Key Technology Research and Application of Responsible Content Recommendation." The team has been dedicated to enhancing the precision, explainability, and fairness of recommendation systems. Key contributions include: - **Multi-scale Knowledge Modeling and Unsampled Learning Recommendation**: The team developed methods to improve recommendation accuracy by integrating intent recognition, knowledge enhancement, and training optimization. - **Explainable Recommendation**: They pioneered the concept of "explainable recommendation," focusing on making both the recommendation outcomes and the reasoning processes transparent and understandable, which has gained international recognition. - **Fair Recommendation**: The project addressed fairness in recommendation systems, proposing innovative methods to ensure equitable treatment for users and content providers. These advancements have significantly mitigated issues of bias and unfairness in recommendation algorithms. The project has produced numerous cutting-edge research outcomes, receiving extensive positive citations from both domestic and international scholars. In the past five years, the research has been cited over 10,000 times. Additionally, the team won the SIGIR 2024 Test of Time Award, marking the first time a mainland Chinese institution has received this accolade, along with six best paper awards and highly cited paper honors from major international conferences. The technology serves over 700 million users and has generated substantial social and economic benefits. **2. Legal Intelligence: Technical Invention First Prize** Professor Maosong Sun's team, in partnership with Beijing Merit Intelligent Technology Co., Ltd. and Professor Weixing Shen from Tsinghua University's Law School, was awarded the Technical Invention First Prize for their project "Key Technology and Application of Deep Integration of Data-Driven and Knowledge-Guided Legal Intelligence." The project, initiated in 2016, is supported by the National 973 Program and the National Key Research and Development Program. It focuses on the integration of artificial intelligence and legal systems to address critical real-world needs such as judicial decision-making and contract review. Key achievements include: - **Data-Driven and Knowledge-Guided Legal Intelligence System**: The team developed a comprehensive legal intelligence system that combines deep learning techniques with expert legal knowledge, ensuring full intellectual property control. - **Applications and Impact**: The "Merit" contract intelligent review system, a product of this project, has become a leading solution in the domestic market. It has been widely applied, contributing to the automation and accuracy of contract reviews and judicial processes. - **Talent Development**: The project has also fostered a new generation of interdisciplinary experts, blending computer science and legal knowledge. The project has generated significant social and economic benefits, with the technology being used in various legal applications and enhancing the efficiency and reliability of legal services. **3. Open-Scene Natural Language Generation: Technical Invention First Prize** Professor Minlie Huang's team, working with Beijing Century Tal Education Technology Co., Ltd., won the Technical Invention First Prize for their project "Key Technology and Application of Open-Scene Natural Language Generation." The team tackled core challenges in natural language generation, including robustness, efficiency, long text generation, and quality evaluation. Key advancements include: - **Non-Autoregressive Generation Models**: The team introduced a novel theory for efficient non-autoregressive generation and developed a directed acyclic graph (DAG) modeling method, surpassing the quality of autoregressive models while achieving a 7-14 times speedup. - **Long Text Generation Methods**: They conducted extensive research on knowledge representation and planning for long text generation, earning the Field Chair Award at ACL 2023 and multiple highly cited paper honors. - **Comprehensive Evaluation System**: The project established a robust evaluation framework for natural language generation, covering data, methods, and platforms. This system has been adopted by over 70 renowned research institutions, significantly impacting the academic community. The technology has been integrated into the Chinese essay automatic evaluation system developed by Tal Education, with over 1 billion cumulative calls in the past three years, serving millions of middle and high school students. This application has not only improved educational outcomes but also generated notable social and economic benefits. **About the Qian Weichang Chinese Information Processing Science and Technology Award** The Qian Weichang Chinese Information Processing Science and Technology Award is the highest scientific and technological honor in the field of Chinese information processing. It recognizes projects that have made original innovations or significant breakthroughs in fundamental methods or key technologies, contributing to the advancement of Chinese information processing and generating substantial economic or social benefits. In 2024, six projects were awarded, with two each in the Natural Science First Prize, Technical Invention First Prize, and Technological Progress First Prize categories. **Conclusion** The recognition of these three projects from Tsinghua University's Computer Science Department underscores the institution's commitment to advancing the field of Chinese information processing through innovative research and practical applications. The awards highlight the importance of interdisciplinary collaboration and the potential for technology to address real-world challenges in content recommendation, legal intelligence, and natural language generation.

Related Links