Research Fellow (Department of Information Systems And Analytics)

17 Mar 2023
End of advertisement period
16 Apr 2023
Contract Type
Full Time

Job Description

In recent times, the Internet has become a popular source for people to seek information. For healthcare information, people often visit online forums and browse medical articles to find answers to their questions about ailments, symptoms, and remedies. However, much of the content in online health forums is user-generated and may misguide users due to inaccuracies, failure to update the information, and lack of validation by healthcare professionals. Thus, there is an unmet need for personal healthcare question-answer (QA) assistants, which refer to software that can glean the required information from online knowledge resources and user’s personal details to present tailored answers to users’ questions.

A key component of such QA assistants are QA systems, which are considered vital tools to help address healthcare information needs of people. A QA system refers to a software that provides natural language responses to user queries by processing unstructured online text and/or by using a structured knowledge base. While progress has been made in QA techniques, QA systems continue to face challenges in their design, integration, and performance. These issues are being tackled by leveraging on advances in the field of natural language processing (NLP), particularly knowledge graphs and language models.

Motivated thus, this project aims to address the research question: What are the design steps and evaluation means for integrating KG and LM approaches for a healthcare QA system? To this end, our approach entails the design iterations of: (i) developing a disease-specific KG for the QA system, (ii) fine-tuning LMs with annotated data, and (iii) joint reasoning over KG and LM for the system. While we intend to formulate and evaluate our design steps for a QA system for a single disease to start with, these are extendable to other diseases. We will gather research articles on specific disease from reliable medical sources to create our KG. We will then use question-context-answer sets to fine-tune popular LMs in this area. Preliminary investigations suggest that this approach has the potential to perform better in terms of providing accurate answers for long sentence questions than using only a KG based approach.

The Research Fellow will be a part of the above project. He/she will be responsible for undertaking in-depth research and innovation in this area that leads to real-world applications as well as publishing the work in top-tier international conferences and journals. He/she will be working closely with the Principal Investigator and with collaborators on interesting and challenging problems in this focal area.


  • PhD degree in a relevant discipline
  • Research areas related to AI, collaboration in domains such as customer service, healthcare
  • Machine learning
  • Econometric analyses - desirable
  • Writing research papers in English
  • Good communication and teamwork skills

Covid-19 Message

At NUS, the health and safety of our staff and students are one of our utmost priorities, and COVID-vaccination supports our commitment to ensure the safety of our community and to make NUS as safe and welcoming as possible. Many of our roles require a significant amount of physical interactions with students/staff/public members. Even for job roles that may be performed remotely, there will be instances where on-campus presence is required.

Taking into consideration the health and well-being of our staff and students and to better protect everyone in the campus, applicants are strongly encouraged to have themselves fully COVID-19 vaccinated to secure successful employment with NUS.

More Information

Location: Kent Ridge Campus
Organization: School of Computing
Department : Department of Information Systems And Analytics
Employee Referral Eligible: No
Job requisition ID : 18950