About
Intro | Research | Funding | Teaching | PhD Thesis
Introduction
Dr. Bart Gajderowicz is a computer scientist at the University of Toronto, with applications in computational social sciences and knowledge representation. His research interests focus on artificial intelligence (AI) methods for modelling, simulating, and evaluating complex social systems as well as the computer systems that support them. His work in these areas combines his passion for technology, healthy communities, and collaboration across sectors. He is an expert in AI systems for smart cities and impact measurements for sustainable communities. The long-term goal of his research is the application of AI techniques towards understanding community behaviour in a smart city context through data-driven decision making.
Dr. Gajderowicz is a Senior Research Associate at Industrial Engineering at the University of Toronto and serves as the Executive Director at the Urban Data Centre (UDC), School of Cities, University of Toronto. His extensive research contributions and affiliations demonstrate his dedication to advancing the field of artificial intelligence (AI) and its application in social services and improving societal well-being through innovative AI applications. His expertise in emulating agent behaviour, ontology matching, and AI-driven planning algorithms has led to significant advancements in evaluating social services and their impact on service clients. Dr. Gajderowicz’s current research focuses on automating impact model extraction from text, service plan generation, and plan evaluation based on the theory of needs/satisfiers, decision-making models, and developing service-related ontologies. His expertise extends to ontology matching, data modelling, and natural language understanding algorithms. He has received multiple fellowships, awards, and funding for his research, demonstrating his contributions to the field. He has applied for and received (as the sole applicant or co-applicant) $295,000 in direct research funding, with an additional $138,000 in grants throughout graduate school.
Current Research
Currently, he serves as the Director of the SeMantIc roLe Extraction project (SMILE), which focuses on explainable natural language understanding AI models for extracting Impact Model information from text. The project was funded by a Mitacs grant worth $25,000 for one year, of which he was the co-applicant and received 100% of the funding. His involvement included the development of the Scroll architecture, an integral part of the first version of SMILE. Dr. Gajderowicz’s work contributes to developing an ontology and methodology for dynamic needs/satisfier planning, which was incorporated into the Compass platform, used by the proposed system’s industry partner, Help Seeker Technologies Inc. As a research fellow at the University of Toronto’s Centre for Social Service Engineering, he develops service-focused ontologies, data standards, artificial intelligence language models, and knowledge extraction methods. He supervises two to six undergraduate and graduate students each semester. At the UDC, he manages the Urban Data Repository and Catalogue project, where his team develops tools for cataloging datasets about urban centres worldwide. In this role, he supervises two to three undergraduate and graduate students each semester. As a co-author, Dr. Gajderowicz actively contributes to the Common Impact Data Standard (CIDS), an ontological basis for extracting Impact Model information, and the Compass Ontology, an extension of the standard used in the HelpSeeker system. His adaPtive grAph Repository pipeLine for dynAmic kNowledge sourCEs (PARLANCE) project involves developing data translation and consolidation logic, enabling the Compass platform to merge diverse data sources into a single representation stored in a knowledge graph.
Bart Gajderowicz has significantly contributed to the proposed project in the following areas. He is the director of SeMantIc roLe Extraction project (SMILE) and developed the initial version being used by HelpSeeker Inc today. He has worked extensively evaluating and contributing to recent advances in hybrid-AI methods that combine symbolic and sub-symbolic methods, used extensively in composite AI models. Specifically, he has extensive experience building Blackboard architectures used by the current version of SMILE. He has evaluated and used a number of Large Language Models (LLMs) for a variety of NLU tasks, many of them integrated as knowledge sources for the SMILE Blackboard architecture, including text generation, named entity recognition, topic modelling, sentiment analysis, question-answering, summarization, and text neutralization (i.e. anti-bias text analysis). Furthermore, he co-authored the CIDS standard and Compass Ontology, a CIDS extension for HelpSeeker, which serves as the ontological basis for extracting Impact Model information in SMILE. He is the lead researcher and developer of PARLANCE, a data translation and consolidation library that enables the merging of various data sources into a knowledge graph, specifically the Compass Knowledge Graph. Additionally, he is a founding member and serves on the Organizing Committee of the Ontologies for Services and Society Workshops (OSS2022 and OSS2023). These workshops aim to enhance interdisciplinary collaboration and communication in the realm of semantic technologies, society, and services in the smart city context. He is responsible for organizing, reviewing, and managing submissions and overseeing outreach efforts.
Teaching
Please see my full Teaching experience
Past Funding and Positions
Dr. Gajderowicz initiated the SMILE project in partneship with HelpSeeker, focusing on explainable natural language understanding AI models. The project was funded by a Mitacs grant worth $25,000 for one year. Additionally, Dr. Gajderowicz held a Mitacs postdoctoral fellow position at Lakehead University, where he collaborates with Dr. Vijay Mago and Wondeur AI, an industry partner, and was awarded a $210,000 Mitacs grant over two years, receiving 100% of the funding. This project extended social simulation models developed during his Ph.D. thesis to capture the behaviour of the art world, which heavily relies on subjective metrics, social norms, and market trends. His responsibilities included developing the simulation research program and incorporating social indicators. He collaborates with fellow researchers, devises publication strategies, and mentors junior machine learning engineers. Simultaneously, at Lakehead University’s DataLab, he supervised graduate students and contributed to applied research in social and economic problems using machine learning, natural language processing, and simulation. This work resulted in a journal publication, a patent submission currently in review, and several works in progress based on the submitted patent. He was also involved in research projects as a postdoctoral fellow at Tata Consultancy Services (TCS) and the University of Toronto on a $40,000 grant with 60% designated to his stipend. His contributions included developing simulation frameworks and exploring hybrid decision-making models for the efficient movement of goal-driven agents.
PhD Thesis and Committee
His Ph.D. thesis, completed at the Industrial Engineering Department at the University of Toronto from 2013 to 2019 under the supervision of Professors Mark S. Fox and Michael Grüninger, aimed to emulate the behaviour and predict the progress of social service clients. His efforts created a reasoning framework, the Bounded Rational Agent MotivAtions simulation framework (BRAMA), that emulates human behaviour, where humans act as rational but emotional agents. He developed a service client agent and simulation environment for holistic evaluation of service provisioning from the client’s perspective toward meeting their needs. This work included creating a new ontology, the Ontology of Social Service Needs (OSSN). These contributions were instrumental in developing a high-fidelity simulation environment for evaluating social service policies and addressing the needs of clients. He partnered with the Calgary Homeless Foundation for empirical validation of this work. For his master’s thesis in the Computer Science Department at Toronto Metropolitan University (2008-2011), Dr. Gajderowicz focused on ontology matching and data consolidation, proposing a bottom-up technique using another artificial intelligence method, decision trees, to create conceptual anchors found in data. This approach facilitated the reuse of ontologies across domains, addressing ontology overlap and generalizing ontologies developed for limited use, as plagued by many existing ontologies.
PhD Committee
Prof. Mark S. Fox (Supervisor): Mark S. Fox is a Professor of Industrial Engineering with a cross-appointment in the Department of Computer Science, and a Senior Fellow in the Global Cities Institute at the University of Toronto. Prof. Fox is head of the Enterprise Integration Laboratory and past holder of the NSERC Industrial Research Chair in Enterprise Integration.
Dr. Michael Grüninger (Supervisor): Michael Grüninger is a Professor at the University of Toronto. He returned to Canada after spending five years as an Assistant Research Scientist in the Institute for Systems Research at the University of Maryland College Park and also a Guest Researcher at the National Institute for Standards and Technology (NIST).
Prof. Marion Bogo: Marion Bogo teaches direct clinical social work practice and the theory and practice of social work education. Her research interests focus primarily on competency for professional practice including social work education and clinical social work supervision. In her research, she has developed and tested field education models and innovative approaches to assessment of student and practitioner competence.
Dr. Vicky Stergiopoulos: Vicky Stergiopoulos is a clinician scientist and the Psychiatrist-In-Chief at St. Michaels Hospital. She is also the Director of the Division of Adult Psychiatry and Health Systems in the Department of Psychiatry at the University of Toronto. Her clinical, education and research activities focus on the development and evaluation of community-based interventions to address the needs of people who are homeless and of those who are frequent users of mental health services. Dr. Stergiopoulos has a keen interest in mental health policy, community development and the redesign of our system of mental health care for the purpose of system improvement. She has worked towards hospital community integration and system level coordination through collaborative community-based clinical program development and research. She is a member of the Board of Directors of the Inner City Family Health Team and Inner City Health Associates. She has held a MOHLTC Career Scientist Award, a Canadian Association for Medical Education Award, a Recognition Award by the Ontario College of Family Physicians as well as the Henry Durost Award for Excellence in Creative Professional Activity in the Department of Psychiatry.