About

I am Ziyan Han, currently a Visiting Scholar at Shenzhen University. I received my Ph.D. at Beihang University (BUAA), under the supervision of Prof. Wenfei Fan and Dr. Yaoshu Wang. Before that, I received the B.E. degree at Xidian University (XDU). My research interests include data management, data quality, data mining, data cleaning, rule discovery, and the intersection of DB and AI.

I was recognized as a Distinguished Graduate of Beihang University (2025) and have been awarded the SIGMOD Student Travel Grant three times (2023–2025).

News

• May 2024 Our paper Discovering Top-k Relevant and Diversified Rules was accepted to SIGMOD 2025.
• Oct 2023 Our paper Splitting Tuples of Mismatched Entities was accepted to SIGMOD 2024.
• Nov 2022 Our paper Discovering Top-k Rules using Subjective and Objective Criteria was accepted to SIGMOD 2023.
• Mar 2022 Our paper Parallel Rule Discovery from Large Datasets by Sampling was accepted to SIGMOD 2022.
• Mar 2021 I joined Shenzhen Institute of Computing Science (SICS) as a research intern.
• Dec 2019 Our paper Keys as features for graph entity matching was accepted to ICDE 2020.

Research Interests

My research primarily focuses on data quality, rule discovery, and data cleaning, with publications in top-tier database conferences, such as SIGMOD and ICDE. Below is a concise overview of my past work.

Data Mining and Data Analysis

I have tackled several challenges in rule discovery, including high computational costs and extensive resource consumption [SIGMOD22], the limitations of non-comprehensive rule evaluation metrics that lack subjective criteria [SIGMOD23], and redundancy within mined rule sets [SIGMOD25].

Data Quality and Data Cleaning

I have developed methods for resolving conflicts within tuples of mismatched entities [SIGMOD24], and for graph entity resolution using graph keys [ICDE20].

Logic Deduction combined with Machine Learning Models

I have integrated machine learning techniques with logic rules to enhance data quality. Specifically, I utilize machine learning techniques to accelerate the rule discovery process [SIGMOD22, SIGMOD23, SIGMOD25]. Additionally, rules discovered can be further applied to improve data quality, such as entity resolution, conflict resolution, and tuple splitting [SIGMOD24, ICDE20].