The 2024 ACM SIGMOD/PODS Conference, a prestigious international event in the database field, was recently held in Santiago, Chile.
During the conference, two teams comprising of undergraduate and postgraduate students from the Southern University of Science and Technology (SUSTech) and Zhejiang University, Alaya and biejuanle, participated in the 2024 SIGMOD Programming Contest. Among them, the Alaya emerged as the winner of the competition.
The participating teams were mentored by Associate Professor Bo TANG from the Department of Computer Science and Engineering (CSE) at SUSTech and Professor Huan LI from Zhejiang University.
The Alaya team included SUSTech undergraduate students Yujun HE, Yitao ZHENG, and Yanqi CHEN from the Department of CSE, along with graduate student Weijian CHEN, doctoral student Long XIANG, and undergraduate students Bowen ZENG and Yu LEI from Zhejiang University.
The second team, biejuanle, consisted of SUSTech undergraduate students Chaoyang HONG, Wanting LI, Zhaohang FENG, Peiran LIANG, Jiale ZHANG, and Yujie WANG, along with undergraduate student Hao WU from Zhejiang University.
This marks the third time that SUSTech’s Database Research Group, part of the Department of CSE, has secured a top prize since it began organizing student participation in the competition in 2020.
The competition’s challenge involved constructing and querying a vector retrieval index under attribute constraints. Participants were provided with 10 million 100-dimensional vector data points encoded by Microsoft’s Turing v5, a large-scale natural language representation model. The teams were required to index the data within a specified time and complete four types of retrieval tasks.
The complexity of vector retrieval under attribute constraints presented significant challenges to traditional graph-based vector index structures. The SUSTech students proposed innovative solutions that delivered impressive results in terms of both timeliness and accuracy. Notably, despite using approximate algorithms, their solutions achieved a recall rate close to 100%, exceeding the minimum margin of error recognized by the evaluation system.
The ACM SIGMOD is part of the Association for Computing Machinery’s (ACM) SIG conference series, which began in 1970. It is widely regarded as the leading international conference in data management, databases, and data science.
The annual SIGMOD Programming Contest addresses various real-world data management challenges and aims to promote academic exchanges among postgraduate students in data science from universities and research institutes worldwide, enhancing their practical problem-solving skills.
Established in 2017 by Associate Professor Bo TANG, the Database Research Group at SUSTech covers the full spectrum of data processing technologies, including database systems, data query processing algorithms, and data visualization.
