Mengzhao Wang

PhD candidate

Zhejiang University

About

I am currently a second-year Ph.D. student in the College of Computer Science and Technology at Zhejiang University. My research interests include high-dimensional data storage, retrieval, and their applications in large language models (LLMs).

My current research focuses on the following topics:

Graph-based vector indexing and retrieval algorithms
Complex query processing in high-dimensional data
Vector database and its applications in large language models

News

[2024.07] Prof. Phil Bernstein remarked that our work is very interesting!
[2024.05] I received the ACM SIGMOD Student Travel Award!
[2024.05] Our demo work is accepted at VLDB 2024!
[2023.11] Our research work is accepted at SIGMOD 2024!
[2023.11] Our research work is accepted at ICDE 2024!
[2023.09] Our research work is accepted at NeurIPS 2023!
[2022.03] Our research work is released in arXiv!
[2021.07] Our research work is accepted by Knowledge-Based Systems!
[2021.06] One survey work is accepted at VLDB 2021!

Recent Publications

Mengzhao Wang, Haotian Wu, Xiangyu Ke, Yunjun Gao, Xiaoliang Xu, Lu Chen (2024). An Interactive Multi-modal Query Answering System with Retrieval-Augmented Large Language Models. In VLDB 2024 (Demo).

Cite Code Dataset Video

Mengzhao Wang, Weizhi Xu, Xiaomeng Yi, Songlin Wu, Zhangyang Peng, Xiangyu Ke, Yunjun Gao, Xiaoliang Xu, Rentong Guo, Charles Xie (2024). Starling: An I/O-Efficient Disk-Resident Graph Index Framework for High-Dimensional Vector Similarity Search on Data Segment. In SIGMOD 2024.

PDF Cite Code Dataset

Mengzhao Wang, Xiangyu Ke, Xiaoliang Xu, Lu Chen, Yunjun Gao, Pinpin Huang, Runkai Zhu (2024). MUST: An Effective and Scalable Framework for Multimodal Search of Target Modality. In ICDE 2024.

PDF Cite Code Dataset

Mengzhao Wang, Lingwei Lv, Xiaoliang Xu, Yuxiang Wang, Qiang Yue, Jiongkang Ni (2023). An Efficient and Robust Framework for Approximate Nearest Neighbor Search with Attribute Constraint. In NeurIPS 2023.

PDF Cite Code Dataset Poster

Mengzhao Wang, Lingwei Lv, Xiaoliang Xu, Yuxiang Wang, Qiang Yue, Jiongkang Ni (2022). Navigable Proximity Graph-Driven Native Hybrid Queries with Structured and Unstructured Constraints. In arXiv 2022.

PDF Cite Code Dataset

See all publications

Invited Talks

Graph-Based Vector Search: Index Layout and Search Optimizations on Disk, ByteDance, 2024.08.
Vector Similarity Search: Disk-Based Graph Index Layout and Search Optimizations, Huawei (Data Storage Technology Planning Dept, Nanjing), 2024.08.
Vector Databases: Index Layout and Search Optimizations on SSD, Baidu, 2024.06.
High-Dimensional Vector Similarity Search, Huawei (Hangzhou Research Inst.), 2023.12.
Approximate Nearest Neighbor Search with Attribute Constraint, Huawei (MRC, Germany), 2023.11.
Research on Large-Scale High-Dimensional Vector Processing, Alibaba DAMO Academy, 2023.11.
Graph-Based Multi-Vector Search and GPU Acceleration, Datawhale, 2023.03.
Graph-Based Approximate Nearest Neighbor Search: An Experimental Study, Zilliz, 2021.09.

Awards

SIGMOD Student Travel Award, 2024.
SIGMOD Student Programming Contest Finalist, 2024.
Outstanding Master’s Thesis of Zhejiang Province, Hangzhou Dianzi University, 2023.
Outstanding Graduate of Zhejiang Province, Hangzhou Dianzi University, 2022.
National Scholarship for Graduate Student, Hangzhou Dianzi University, 2021.
First Prize of Academic Scholarship, Hangzhou Dianzi University, 2021.
First Prize of Academic Scholarship, Hangzhou Dianzi University, 2020.

Experience

2022.09-Now: PhD candidate in College of Computer Science and Technology, Zhejiang University. Supervisor: Prof. Yunjun Gao(Homepage, Scholar)
2022.05-2022.08: Research Engineer Intern in Research Department, Zilliz.
2021.09-2022.09: R&D Engineer Intern in Big Data Intelligence Research Center, Zhejiang Lab.
2019.09-2022.03: M.E. in School of Computer Science and Technology, Hangzhou Dianzi University. Supervisor: Prof. Xiaoliang Xu(Homepage, Scholar)
2015.09-2019.06: B.E. in College of Materials Science and Engineering, Qingdao University of Science and Technology. Supervisor: Dr. Yingjie Chen(Homepage, Scholar)

Selected Comments

2024-07: Prof. Phil Bernstein remarked in an email that our paper “Starling: An I/O-Efficient Disk-Resident Graph Index Framework for High-Dimensional Vector Similarity Search on Data Segment” presented at SIGMOD 2024 is very interesting.
2024-07: Prof. Dixin Tang expressed interest in our work “Starling: An I/O-Efficient Disk-Resident Graph Index Framework for High-Dimensional Vector Similarity Search on Data Segment” and requested more details via email.
2024-07: Prof. Jieming Shi showed interest in our work “Starling: An I/O-Efficient Disk-Resident Graph Index Framework for High-Dimensional Vector Similarity Search on Data Segment” and asked for more information via WeChat.

Contact

wmzssy@zju.edu.cn/wmengzhao@qq.com
38 Zheda Road, Hangzhou, Xihu District 310027
Add Me on Wechat by Scanning the QR Code
Discuss on CSDN