A Survey on AI-based Table Column Semantic Type Annotation Methods

PhD Qualifying Examination


Title: "A Survey on AI-based Table Column Semantic Type Annotation Methods"

by

Mr. Yushi SUN


Abstract:

Table column semantic type annotation is crucial for a variety of data 
preparation and discovery tasks within the database domain. Accurately 
annotating the data types of table columns holds tremendous benefits across a 
diverse range of applications. Specifically, precise annotation of column type 
enhances the quality of data integration, schema matching, data cleaning, and 
data visualization. Due to its great value and importance in the data 
preparation and integration field, there is a growing interest in developing 
advanced table column semantic type annotation methods. Recent methods 
introduce state-of-the-art AI techniques for this task to achieve improvement 
in annotation quality. Despite the rapid advancement in the performance of 
AI-based table column semantic type annotation approaches, there is a lack of a 
systematic review of existing approaches to summarize the limitations of 
existing solutions, challenges that remain unsolved, and the future directions 
of this task.

In this survey, we first introduce the background of the table column semantic 
type annotation task. We then define the important concepts and the problems of 
the task. Based on the problem statement and the concepts defined, we classify 
the existing AI-based table column semantic type annotation methods and 
evaluate the advantages and disadvantages of each category. The survey ends 
with a discussion of the future directions and opportunities based on the 
classification and evaluation of the existing approaches.


Date:                   Friday, 12 July 2024

Time:                   2:00pm - 4:00pm

Venue:                  Room 3494
                        Lifts 25/26

Committee Members:      Prof. Lei Chen (Supervisor)
                        Prof. Xiaofang Zhou (Chairperson)
                        Dr. Dan Xu
                        Dr. Lei Li (HKUST-GZ)