Overview of NUS's Pioneering AI-Humanities Fusion
The National University of Singapore (NUS), consistently ranked among Asia's top universities, has launched the Centre for Computational Social Science and Humanities (CSSH) to bridge artificial intelligence (AI) with humanities and social sciences. Established in July 2024 and officially unveiled on March 4, 2026, this interdisciplinary hub aims to harness computational tools for deeper insights into human behavior, culture, and society.
In Singapore's vibrant higher education landscape, where NUS and Nanyang Technological University (NTU) lead global AI and computer science rankings—NUS at #13 and NTU at #16 in THE 2026—CSSH exemplifies the nation's push toward AI-driven social research.
Leadership Driving Interdisciplinary Innovation
Professor Atreyi Kankanhalli, a renowned expert in intelligent systems and human-centered AI, brings her experience from projects like "The Other Me," which develops personalized AI assistants for daily life.
Together, they oversee more than 50 projects, providing seed funding to spark innovative ideas that scale into major grants. This leadership ensures CSSH not only advances research but also trains the next generation of scholars at the intersection of AI and humanities—a critical skill set for Singapore's faculty positions in emerging fields.
Core Research Themes and Methodologies
CSSH organizes its work around three pillars: AI and computational models of human behavior, health and social care, and texts, heritage, and culture. Computational social science involves using AI techniques—such as machine learning (ML), large language models (LLMs), and network analysis—to process big data from digital traces like tweets, forums, and digitized texts.
Step-by-step, researchers collect data (e.g., social media APIs), preprocess it (cleaning noise, tokenization), apply models (LLMs for sentiment analysis), and validate with traditional methods like surveys. This hybrid approach uncovers patterns invisible to manual analysis, such as evolving public sentiments or cultural shifts.
- Human behavior modeling: Predicting responses via simulations.
- Health/social care: Analyzing addiction trends from online forums.
- Heritage: Digitizing and interpreting historical documents.
In Singapore, where 91% of residents are digitally connected, this yields rich datasets for real-world impact.
Flagship Project: Simulating Public Opinion for Policy Design
The Computational Social Simulations for Aiding Policy Design, a five-year SSRC-funded initiative, exemplifies CSSH's policy relevance. Led by 14 principal investigators across NUS, NTU, SMU, and SUTD, it uses LLMs to simulate societal reactions to policies on heritage conservation, sustainability, and health.
Process: (1) Input policy scenarios into LLMs trained on diverse texts; (2) Generate synthetic responses mimicking demographics; (3) Analyze for polarization or support; (4) Validate with real surveys, especially for underrepresented groups like the elderly. Prof Kankanhalli notes this complements costly field studies (hundreds of thousands of dollars), enabling rapid iteration.
This could transform policymaking, reducing risks in Singapore's data-rich environment. For aspiring researchers, such projects offer hands-on experience—check research assistant jobs at NUS.
Photo by Roaming Pictures on Unsplash
Preserving Heritage: The Jawi AI Initiative
Led by Asst Prof Escobar Varela with the National Library Board (NLB), Jawi AI tackles pre-1970s Malay-language newspapers in Jawi script—Arabic-based Malay cursive. Using optical character recognition (OCR) fine-tuned on historical fonts, it converts scanned pages into searchable modern Malay text.
Timeline: Started June 2025, completion August 2026. Thousands of pages digitized, enabling queries on Singapore's history—family stories, neighborhoods. Future expansions to Mandarin and Tamil support multilingual heritage.
This project underscores AI's role in cultural preservation amid digitization trends, vital for Singapore's multicultural identity.
Analyzing Societal Shifts: Divorce Judgments Study
A 2025 seed-funded project analyzes over one million Chinese first-instance divorce judgments using LLMs. It maps motive evolution (e.g., infidelity to incompatibility) over 40 years, linking to age, education, GDP, urbanization.
Involving NUS Centre for Family and Population Research and Communications departments, it reveals assortative mating patterns. Plans to adapt for Singapore data address local trends like delayed marriages.
Broader Portfolio and Seed Funding Impact
CSSH's 50+ projects include AI-powered clinic notes and more, funded via seed grants. The 2025 cohort spurred collaborations, leading to SSRC scales.Explore seed grants. This ecosystem fosters innovation, with outputs like publications and tools.
Singapore's AI ecosystem—$1B+ invested via AI Singapore—amplifies this, positioning unis as hubs.
Collaborations and Singapore's AI-SSH Ecosystem
CSSH partners with NTU, SMU, NLB, NHB, NUHS. Similar efforts: SMU-A*Star lab (2022) on ageing/polarization. Globally, comp social science booms (e.g., Stanford's centers), but Singapore leads Asia with ethics focus.
Challenges: AI biases in training data; CSSH emphasizes validation and ethics.
Photo by Bing Hui Yau on Unsplash
Implications for Higher Education and Careers
CSSH signals demand for hybrid skills—AI + humanities. NUS eyes undergrad programs, boosting lecturer jobs. Graduates excel in policy, tech, academia; Singapore's AI jobs grew 20% yearly.
For students: Rate My Professor for NUS faculty like Kankanhalli.
Future Outlook: Scaling Impact and Education
With MOE's SSH boost, CSSH will expand simulations, heritage tools, behavior models. Potential undergrad tracks prepare for AI ethics roles. As Singapore aims net-zero AI governance, CSSH leads balanced innovation.
Explore opportunities at higher ed jobs, university jobs, and career advice. Stay tuned via CSSH site and Straits Times.