Deep learning techniques for entity resolution

PhD Qualifying Examination


Title: "Deep learning techniques for entity resolution"

by

Miss Shiwen WU


Abstract:

Entity Resolution (ER) is a critical data processing task that determines 
whether two data entries refer to the same real-world entity. This process 
typically comprises two stages: blocking and matching. ER has attracted 
substantial attention, leading to the development of a broad range of 
methodologies, including rule-based approaches, traditional machine learning 
techniques, and advanced deep learning frameworks. Notably, deep learning 
techniques excel at capturing latent semantic patterns and contextual 
information in data, thereby achieving state-of-the-art results in ER tasks. 
Given the pivotal role of deep learning in this field, this article aims to 
provide a comprehensive review of recent advancements in deep learning-based 
(DL-based) ER frameworks. We introduce a taxonomy to classify existing works 
based on their learning paradigms and processing steps, showcasing 
representative studies and analyzing their strengths and limitations. 
Furthermore, this article offers new perspectives and suggests directions for 
future research in ER.


Date:                   Monday, 3 June 2024

Time:                   10:00am - 12:00noon

Venue:                  Room 3494
                        Lifts 25/26

Committee Members:      Prof. Xiaofang Zhou (Supervisor)
                        Prof. Raymond Wong (Chairperson)
                        Dr. Wilfred Ng
                        Dr. Wen Hua (PolyU)