In order to facilitate software interaction and increase user satisfaction, various research efforts have tackled the problem of software customization by modeling the user’s goals, skills, and preferences. In this thesis, we focus on run-time solutions for adapting various interface and interaction aspects of software. From an intelligent agent’s perspective, the system views this customization problem as a decision-theoretic planning problem under uncertainty about the user. We propose a methodological framework for developing intelligent software interaction and assistance. This framework has been instantiated in various case studies which are reviewed in the thesis. Through efforts of data collection experiments to learn model parameters, simulation experiments to assess system feasibility and adaptivity, and usability testing to assess user receptiveness, our case studies show that our approach can effectively carry out customizations according to different user preferences and adapt to changing preferences over time.
Identifer | oai:union.ndltd.org:TORONTO/oai:tspace.library.utoronto.ca:1807/31785 |
Date | 09 January 2012 |
Creators | Hui, Bowen |
Contributors | Boutilier, Craig |
Source Sets | University of Toronto |
Language | en_ca |
Detected Language | English |
Type | Thesis |
Page generated in 0.0015 seconds