The COoKIE Group
Communication, Organization of Knowledge, Information Ecosystems
What we Do
The COoKIE is an interdisciplinary Human-Computer Interaction research group focusing on Human-AI communication and collaboration.
Our recent work touches on diverse technologies, including conversational agents, large language models, generative AI systems, and AI-based decision support systems. Seeking to advance Human-AI interaction design, we explore mechanisms driving the perceptions of different types of AI systems, such as trust, anthropomorphization, or perceived reliability to inform the system and algorithm design requirements for effective, efficient, and ethical AI.
Our Work is Supported by:
The Natural Sciences and Engineering Research Council of Canada (NSERC)
The Social Sciences and Humanities Research Council of Canada (SSHRC)
The Naver Corporation,
South Korea
The Faculty of Information,
University of Toronto, Canada
Current Projects
Human-Agent Communication
The growing complexity of intelligent systems creates a promising environment for the shift in a user experience paradigm: from pure human-system interaction to human-system communication and collaboration. This is further amplified by the increased popularity of entity-based system interfaces (artificial agents) and the integration of conversational user interfaces (CUI). The human-agent communication paradigm opens exciting opportunities to exploit a computer's unique abilities to complement humans. However, first, we need to understand the specifics of human-agent communication.
AI-Based Decision Support Systems (ADS)
AI-integrated decision-making support systems allow to augment and extend human capabilities, improving the quality and efficiency of decisions. These systems use AI to learn, remember, analyze, and reason based on large amounts of often complex data. The successful use of these features requires an understanding of how they can appropriately complement human processes. To address these questions, we explore what aspects of augmentation are required in different scenarios, how the outcomes should be communicated and explained (xAI), and how to properly calibrate and, if needed, repair trust in systems' performance.
Information Architecture for Conversational UI
The intuitive nature of Conversational User Interfaces (CUI), fashioned after natural human conversation, allows for convenient and fast input and can serve as a single-channel input for multiple services, devices, applications, and even environments. However, due to the absence of visual clues, affordances, and spatial organization, the use of CUI is associated with a decreased sense of control over the information dynamics. We explore the perceived “locations”, “structure“, and “paths” of the information shared through CUI, to inform the required information architecture.
Meet the Team
Anastasia Kuzminykh
Assistant Professor,
Faculty of Information, UofT
Director of The COoKIE Group and the Toronto Human-AI Interaction Research School (THAI RS)
Nazar Ponochevnyi
Undergrad. Research Intern,
Faculty of Information, UofT
Developing Voice Interface for Chart Creation
Paula Aoyagui
Research Assistant, Previously Master's Student, Faculty of Information, UofT
Explainable AI Strategies in Non-Ground-Truth Cases
Batool Fatima
Master's Reading Student,
Faculty of Information, UofT
Trust Mechanisms in Human-AI Collaboration
Christina Wei
Ph.D. Student, THAI RS'22,
Faculty of Information, UofT
Conversation Architecture for Artificial Agents
Yuxin Xu
Master's Thesis Student,
Faculty of Information, UofT
Image Accessibility for Visually Impaired Users
Rimsha Rizvi
Master's Research Assistant, Reading Student, Faculty of Information, UofT
Fairness Mitigation in Human-AI Collaboration
Omer Imran
Master's Reading Student,
Faculty of Information, UofT
Trust Mechanisms in Human-AI Collaboration
Rezvan Boostani
Ph.D. Student,
Faculty of Information, UofT
Artificial Agents for Financial Assistants for Older Adults
Fatima Zohra
Master's Thesis Student,
Faculty of Information, UofT
Accuracy Preserving Fairness Mitigation Techniques for Financial ML Models
Kshitij Anand
Master's Reading Student, Faculty of Information, UofT
Conversational Information Architecture in Human-AI Collaborative Tasks
Janna Cameron
Master's Reading Student,
Faculty of Information, UofT
Trust Calibration Design Techniques in Human-AI Collaboration
Senjuti Dutta
Ph.D. Student, Electrical Eng. & Computer Science, UTK
Intelligent Systems to Support Crowdwork Flexibility
Janet Lu
Master's Thesis Student,
Faculty of Information, UofT
Conversational Agents for Co-Design Assistance
Shradha Anand
Master's Research Assistant,
Faculty of Information, UofT
Document Exploration Through Conversational Interfaces
Juan Antonio Nelson
Master's Research Assistant,
Faculty of Information, UofT
Automatic Speech Detection with a focus on Caribbean Dialects
Manveer Kalirai
Ph.D. Student, THAI RS'22, Faculty of Information, UofT
Information Architecture for Smart Homes
Nabila Chowdhury
PhD Research Assistant, THAI RS'22, Faculty of Information, UofT
Promt-Based Bias in AI Generated Art
Mohammed Elkhechen
Master's Research Assistant,
Faculty of Information, UofT
Psychological Ownership in Human-AI Collaboration
Nimisha Karnatak
Ph.D. Student, Faculty of Information, UofT
Explainable Medical AI for Doctors in Rural Areas
Sharon Ferguson
PhD Research Assistant, THAI RS'21, Mechanical and Industrial Engineering, UofT
Architecture of AI Explanations of Subtle Sexism
Tanisha Amarakoon
Master's Research Assistant,
Faculty of Information, UofT
Psychological Ownership in Human-AI Collaboration
Alumni
Hidaya Ismail
Research Assistant,
Faculty of Information, UofT
Psychological Ownership in Human-AI Collaboration
Mona Abou Swaid
Master's Research Assistant,
Faculty of Information, UofT
Sense of Ownership Over Digital Artifacts
Deepak Surya
Mitacs Globalink Research Intern, 2022, Faculty of Information, UofT
Text-to-Voice Conversation Style Transformation
Akriti Kaur
Master's Research Assistant,
Faculty of Information, UofT
Document Exploration Through Conversational Interfaces
Kusum Grandhi
Mitacs Globalink Research Intern, 2022, Faculty of Information, UofT
Modality-Dependant Decision Effects of AI Explanations
Lizhen Ying
Master's Reading Student,
Faculty of Information, UofT
Anthropomorphization of Disembodied Agents
Kelly McConvey
Ph.D. Research Assistant,
Faculty of Information, UofT
Collaboration with AI Models in Education
Mujgan Ozceylan
Master's Research Assistant, Reading Student, THAI RS'21
Faculty of Information, UofT
AI Explanations of Subtle Sexism
Lakshya
Master's Reading Student,
Faculty of Information, UofT
Designing Human-AI Conversation
Julia Gil
Undergrad Reading Student,
Faculty of Information, UofT
User Communication with Conversational Agents
Jinchi Lin
Master's Reading Student,
Faculty of Information, UofT
Personalization of Disembodied Intelligent Agents
Ruijia Yang
Undergrad Reading Student,
Computer Science, UofT
Simulation As a
Research Method in HCI
Bill Than
NSERC USRA, BI student,
Faculty of Information, UofT
Engagement Dynamics for Teleconferencing Classrooms During Covid-19
Join the Team
PostDocs
We would be happy to consider applications from potential postdoc candidates. Typically, a postdoc contract is offered for a year with a potential consequent extension for another year. Currently, we are looking for postdoc candidates, broadly interested in Human-AI Collaboration with a particular focus on conversational interactions. If you are interested in a postdoc in our group, please send your CV, writing samples, and a brief statement of research interest to Prof. Anastasia Kuzminykh (anastasia.kuzminykh@utoronto.ca).
Ph.D. Program
We invite applications from candidates with a Master's degree in Computer Science, Systems Design Engineering, Data Science, Psychology, Communication Studies, or similar fields. Research experience is expected from the applicants, academic publications are a plus. The application period is typically open September-November, with the entry into the program once a year, in September. Please review the admission requirements to the Ph.D. in Information Program. Prior to submitting your application, it is highly recommended to contact your potential supervisor and have a discussion about your research plans and interests. If you are interested in a Ph.D. in our group, please send your CV, writing samples, and a brief statement of research interest to Prof. Anastasia Kuzminykh (anastasia.kuzminykh@utoronto.ca).
Master's Thesis
The Master's Thesis option at the COoKIE group is available to students, enrolled in the Master's of Information program at the iSchool, University of Toronto. Please review the pre-requisites and eligibility for the master's thesis. To inquire about the reading course and/or the thesis work at the COoKIE Group, please send your CV and a brief statement of research interest to Prof. Anastasia Kuzminykh (anastasia.kuzminykh@utoronto.ca). If applicable, please mention any relevant experience, including prior research experience and/or reading courses taken in the past.
RAship
In most of the academic terms (September-December, January-April, May-August), the COoKIE Group has 1-2 Research Assistantship positions available for Master's or undergraduate level students. While typically we hire students from the University of Toronto, external applications can occasionally be considered. The standard RA contract in our group includes 12 weeks, 5-6 hours per week, working on an individual or a group project. To inquire about open positions and to apply for RAship, please contact Prof. Anastasia Kuzminykh (anastasia.kuzminykh@utoronto.ca) no later than a week before the beginning of an academic term. Please include your CV and a brief statement of research interest.
