An Integrated Model of Tasks and Uncertainties for Designing Task-aware Search Assistants

Abstract: When users conduct a search, it is usually driven by a task they need to accomplish. The search process can be long, interactive for complex tasks, and involve multiple stages with shifting goals and cognitive focus. Users' search strategies are influenced by their intent, encountered problems, and cognitive state of knowledge at these search stages. However, most existing search systems are designed to optimize one request at a time without considering the larger task at hand, the different stages of the task, or the user's cognitive focus throughout the search session. Although a few descriptive and theoretical models in the literature describe the search process and its associated tasks, research has yet to explore dynamic task characteristics in search personalization. Furthermore, there is a lack of support for users to complete their tasks in an adaptive, dynamic way. To address these issues, this dissertation takes a multi-disciplinary, human-centered, and mixed-method approach to develop a conceptual framework for understanding how different types of tasks trigger specific information needs and that can lead to different strategies for seeking different forms of information and information sources and, in the due process, identify any barriers users face. The dissertation also proposes new computational models to construct unified task representations using underlying search behavioral signals and make existing search and retrieval systems more responsive to users' shifting cognitive focus during the search process by using knowledge gained about users' tasks and problems. Ultimately, the goal is to empower users to make informed decisions about different aspects of their lives by providing relevant information that aligns with their current task state.

Supervisory Committee

Chair: Chirag Shah, Professor, Information School
GSR: David Ribes, Associate Professor, Human Centered Design and Engineering
Member: David Hendry, Associate Professor, Information School
Member: William Howe, Associate Professor, Information School
Member: Ryen White, Affiliate Professor, Information School and General Manager, Microsoft Research