Support the sustainable usage and development of the shared knowledge on CQA platforms

The Hong Kong University of Science and Technology
Department of Computer Science and Engineering


PhD Thesis Defence


Title: "Support the sustainable usage and development of the shared knowledge 
on CQA platforms"

By

Mr. Chengzhong LIU


Abstract:

On Community-Based Question Answering (CQA) platforms like Quora and Zhihu, 
people can post questions and get answers from the community. These platforms 
are a popular way for millions of users to learn and disseminate knowledge 
every day. As the amount of shared knowledge on these platforms increases, it 
is important to study how to enhance the long-term use and growth of such 
knowledge repositories for the CQA communities.

The main goal of this thesis is to explore how to support the digestion, 
contribution, and governance of collective knowledge on CQA platforms. We first 
built PlanHelper, an interactive system that helps CQA users digest existing 
answers more effortlessly, especially when they want to create activity plans. 
We achieved this by designing a Natural Language Processing (NLP) pipeline that 
organizes existing answers and building PlanHelper on top of it. Next, we 
constructed CoArgue, another interactive system that helps CQA users contribute 
more effectively to the ongoing discussion, especially on non-factoid topics. 
To do this, we first designed a chatbot with enhanced social intelligence to 
engage users during the interaction. Based on the work PlanHelper, we then 
improved the NLP pipeline to suggest potential points for users to contribute 
to as well as structuring the existing answers. We integrated the chatbot with 
the pipeline and developed CoArgue. Lastly, we moved our attention to CQA 
platform governance, besides supporting user actions on digestion and 
contribution. We identified possible problems of the co-created knowledge 
structure in CQA platforms especially in the context of the community efforts 
to discuss science related questions. Based on our analysis, we proposed design 
recommendations for the CQA platforms to enable effective, accessible, and 
accurate science sensemaking of the general users.

In summary, this thesis sought to enhance the process of building knowledge on 
CQA platforms from both user actions and platform governance. For future work, 
we intended to study the applicability of the proposed solutions and how the 
Generative AI could improve them.


Date:                   Friday, 14 June 2024

Time:                   4:00pm - 6:00pm

Venue:                  Room 3494
                        Lifts 25/26

Chairman:               Dr. Jia LIU (IEDA)

Committee Members:      Dr. Xiaojuan MA (Supervisor)
                        Prof. Cunsheng DING
                        Dr. Shuai WANG
                        Prof. Xun WU (PPOL)
                        Dr. Yan Tina LUXIMON (PolyU)