Towards Scrutable Decision Tree-based User Model utilising Interactive and Interpretable Machine Learning (SUM-IML)
Year: 2021
Type: dissertation
Abstract: Personalising user models has gained considerable attention in\nrecent literature. In an information-rich environment, it is crucial not\nonly to provide the information at any time, at any place, and in any\nform, but to also minimise information overload for the user and ease\ntheir ability to access relevant information in their user model.\nSupplying tailored information and providing offerings that suit each\nuser's interests may be enhanced by involving the user.\nRecently, much research has been concerned with employing\nMachine Learning for user modelling. Machine Learning (ML) tries to\nmimic and predict a user s activities, however, it cannot model the\nuser themselves. Users can benefit from involvement in the modelling\nprocess by incorporating their input as considerations in modelling\nthrough employing interactive Machine Learning or human-in-theloop controls. Such user involvement needs the user to understand\nthe model behaviour to ensure their inputs are effective. This can be\nachieved by utilising Machine Learning interpretability techniques.\nThis work proposes an interpretation of the model to the user in order\nto provide the user with better understanding for the model\nbehaviour.\nIn this study, the proposed approach, termed SUM-IML (Scrutable\nUser Modelling using Interactive and Interpretable Machine\nLearning)., implements model scrutability by combining the benefits\nof interactive ML as well as Interpretable ML in user modelling. This\nthesis presents the research question driving this work, a state-ofthe-art review of user modelling, scrutability, interactive,\ninterpretable Machine Learning literature and the evaluation\nmethodologies required. It then presents two related experiments\nthat demonstrate the exploration of research question through their\nresults and the conclusion.
Author: Dima S. Mahmoud
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W3197284507