Predicting public opinion, preserving historical texts: New NUS centre marries humanities with AI
Sign up now: Get tips on how to help your child succeed
NUS' Centre for Computational Social Science and Humanities deputy director, Associate Professor Miguel Escobar Varela (right), with the centre's co-director, Professor Atreyi Kankanhalli.
ST PHOTO: NG SOR LUAN
- NUS' new Centre for Computational Social Science and Humanities (CSSH) launches to tackle complex social problems using data and AI.
- Researchers will create an AI platform simulating public responses to policies and develop AI tools for transliterating historical Jawi newspapers, preserving heritage.
- CSSH provides seed funding for new ideas and leads larger-scale research projects that are carried out in collaboration with the government, industry or other universities.
AI generated
SINGAPORE – An AI-driven platform will soon be created to simulate public responses to local policies, to stress-test and refine ideas before they are rolled out.
Powered by large language models (LLMs) – artificial intelligence systems trained on vast amounts of text to analyse and generate content – the platform will model how parts of society might react to proposals, such as those related to heritage conservation, sustainability and health.
The five-year undertaking is expected to start in July.
Separately, AI tools are being built to transliterate Singapore’s pre-1970s Malay-language newspapers written in Jawi, a script few can now read. Using optical character recognition, the system converts thousands of archived pages into searchable Malay text, preserving a significant part of the nation’s history.
These initiatives are among more than 50 projects under the new Centre for Computational Social Science and Humanities (CSSH) at the National University of Singapore (NUS) that will be officially launched on March 4.
This comes as interdisciplinary research is becoming more important, with increasingly complex challenges facing humanity that require expertise from multiple fields.
The Singapore Management University (SMU) and A*Star had also set up a lab in 2022 which applies AI technology to tackle high-priority national challenges facing Singapore
NUS’ centre brings together 105 researchers across disciplines to tackle complex issues such as inequality, addiction and polarisation. Set up in July 2024, its aim is to blend humanities and social sciences with computational methods.
By tapping large volumes of online data – from social media to consumer behaviour and travel patterns – the projects aim to generate deeper insights to tackle social problems.
Located at NUS’ Kent Ridge campus, the centre’s work spans themes such as AI and computational models of human behaviour, health and social care, as well as texts, heritage and culture.
Professor Atreyi Kankanhalli, co-director of CSSH, said: “We think of the centre as a matchmaker and a facilitator for interdisciplinary work.”
CSSH provides seed funding for new ideas and leads larger-scale research projects that are carried out together with the government, industry or other universities.
The Computational Social Simulations for Aiding Policy Design project will involve 14 principal investigators, eight collaborators and 20 other research staff and students across NUS, Nanyang Technological University, SMU and the Singapore University of Technology and Design.
It is funded by the Ministry of Education’s (MOE) Social Science Research Council (SSRC).
The team will work with the National Heritage Board, the National University Health System and HDB Health District @ Queenstown, looking at how people respond to heritage conservation policies and preventive health care programmes, said Prof Atreyi.
The effort aims to complement large-scale surveys or field studies which can cost hundreds of thousands of dollars to conduct over time, said Prof Atreyi. She added that offline data will also be used to capture the views of underrepresented groups such as the elderly, and the simulation results will be validated with surveys and field studies as needed.
The Jawi AI project led by Associate Professor Miguel Escobar Varela, deputy director of CSSH, was developed together with the National Library Board (NLB). It hopes to expand research into Malay-language journalism.
He said he hopes the project will enable more people to do research, including professional historians and the public who can use the NLB portal to learn about the history of Singapore, their families or neighbourhoods.
“And so far, that’s something we couldn’t have done with the Jawi newspapers.”
The project began in June 2025 and is scheduled to be completed in August 2026.
Prof Escobar Varela hopes the AI tools can be extended to other languages such as Mandarin and Tamil, to record a “multilingual heritage” of Singapore.
Another project which received CSSH’s seed funding in 2025 is analysing more than one million first-instance divorce judgments using LLMs to study the evolution of major divorce motives over the past 40 years.
Researchers will link motives to traits such as age, education and assortative mating – often called “like mating with like”. It will also examine how macro-level factors like regional gross domestic product, urbanisation and gender ideology influence marital dissolution.
The project brings together researchers from NUS’ Centre for Family and Population Research, Department of Information Systems and Analytics, and Department of Communications and New Media. The team is working with China-based data and hopes to scale the research using Singapore data in future.
Beyond producing research, CSSH is also exploring developing a programme for undergraduates.
MOE has given a boost to social science and humanities research, setting aside its third and largest tranche of funding
The SSRC was set up in 2016 to provide direction for social science and humanities research as Singapore matures as a nation, such as in the areas of ageing and social mobility.
