Dr Savitha Sam Abraham
Contact details
About
Overview
Savitha Sam Abraham is a Lecturer and Researcher in Artificial Intelligence in the Department of Computer Science at City St George’s, University of London. Her research focuses on making reliability a key feature of AI systems through two complementary directions: developing rigorous methods to evaluate AI systems after development, and embedding reliability into AI architectures through sound logical reasoning. Her research includes the design and development of benchmarks that enable systematic evaluation of generative models, including language and image models, on reasoning-intensive tasks. She is particularly interested in the neuro-symbolic paradigm as a means of incorporating structured reasoning into AI systems. Her research also encompasses responsible AI, building on previous work on fairness in machine learning. She is keen to explore interdisciplinary applications of AI in areas such as pedagogy, healthcare, and law, where reliability and trustworthy decision-making are particularly important.
Her teaching includes:
2025-2026: INM359 Object Oriented Programming with C++
2026-2027: INM359 Object Oriented Programming with C++
IN3045/INM434 Natural Language Processing
Qualifications
- PhD in Computer Science and Engineering, Indian Institute of Technology Madras, India, January 2013 - June 2020
Employment
- Post-doctoral researcher, University of Adelaide, Australia, September 2023 - August 2025
- Post-doctoral Researcher, Orebro University, Sweden, March 2021 - September 2023
- Research Scientist, NLP, Buddi.AI, India, June 2020 - February 2021
Languages
English (can read, write, speak, understand spoken, peer review), Hindi (can read, write, speak, understand spoken), Malayalam (can read, write, speak, understand spoken, peer review) and Tamil (can speak, understand spoken)
Publications
Publications by category
Conference papers and proceedings (14)
- Sam-Abraham, S., Totis, P., Alirezaie, M. and de Raedt, L. Can Better Solvers Find Better Matches? Assessing Math-LLM Models in Similar Problem Identification. .
- Abraham, S.S., Garg, S. and Dayoub, F. (2025). To Ask or Not to Ask? Detecting Absence of Information in Vision and Language Navigation. 2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) 26 February-6 March.doi:10.1109/wacv61041.2025.00727
- Aregbede, V., Abraham, S.S., Persson, A., Längkvist, M. and Loutfi, A. (2024). Affordance-Based Goal Imagination for Embodied AI Agents. 2024 IEEE International Conference on Development and Learning (ICDL) 20-23 May.doi:10.1109/icdl61372.2024.10644764
- Abraham, S.S., Alirezaie, M. and De Raedt, L. CLEVR-POC: Reasoning-Intensive Visual Question Answering in Partially Observable Environments. .
- Abraham, S.S., P, D. and Sundaram, S.S. Span Detection for Kinematics Word Problems. .doi:10.1007/978-981-99-1645-0_23
- Lindström, A.D. and Abraham, S.S. (2022). CLEVR-Math: A Dataset for Compositional Language, Visual and Mathematical Reasoning. 16th International Workshop on Neural-Symbolic Learning and Reasoning (NeSy) 28-30 September, Windsor, UK.
- P, D. and Abraham, S.S. Representativity Fairness in Clustering. WebSci '20: 12th ACM Conference on Web Science.doi:10.1145/3394231.3397910
- Sundaram, S.S., Deepak, P. and Abraham, S.S. Distributed representations for arithmeticword problems. .
- Abraham, S.S., Deepak, P. and Sundaram, S.S. Fairness in clustering with multiple sensitive attributes. .doi:10.5441/002/edbt.2020.26
- P., D. and Sam Abraham, S. Fair Outlier Detection. .doi:10.1007/978-3-030-62008-0_31
- Abraham, S.S. and Sundaram, S.S. Combining qualitative and quantitative reasoning for solving kinematics word problems. .
- Sundaram, S.S. and Abraham, S.S. Solving Simple Arithmetic Word Problems Precisely with Schemas. .doi:10.1007/978-3-319-92058-0_52
- Abraham, S.S. and Khemani, D. Hybrid of qualitative and quantitative knowledge models for solving physics word problems. .
- Abraham, S.S. and Idicula, S.M. (2012). Comparison of statistical and semantic similarity techniques for paraphrase identification. 2012 International Conference on Data Science & Engineering (ICDSE) 18-20 July.doi:10.1109/icdse.2012.6282314
Journal articles (6)
- Sundaram, S.S., Gurajada, S., Padmanabhan, D., Abraham, S.S. and Fisichella, M. (2024). Does a language model “understand” high school math? A survey of deep learning based word problem solvers. WIREs Data Mining and Knowledge Discovery, 14(4). doi:10.1002/widm.1534
- K., A., P., D., Sam Abraham, S., V. L., L. and P. Gangan, M. (2022). Readers’ affect: predicting and understanding readers’ emotions with deep learning. Journal of Big Data, 9(1). doi:10.1186/s40537-022-00614-2
- P, D. and Abraham, S.S. (2021). FairLOF: Fairness in Outlier Detection. Data Science and Engineering, 6(4), pp. 485-499. doi:10.1007/s41019-021-00169-x
- P., D. and Sam Abraham, S. (2020). Correction to: Fair Outlier Detection. pp. C1-C1. doi:10.1007/978-3-030-62008-0_42
- Sundaram, S.S. and Abraham, S.S. (2019). Semantic Representation for Age Word Problems with Schemas. New Generation Computing, 37(4), pp. 429-452. doi:10.1007/s00354-019-00069-9
- Abraham, S.S. and Sundaram, S.S. (2019). An Ontology-Based Kinematics Problem Solver Using Qualitative and Quantitative Knowledge. New Generation Computing, 37(4), pp. 551-584. doi:10.1007/s00354-019-00067-x