Posted: 19th June 2017

Oxford student Oana Tifrea-Marciuska has been awarded the EPSRC Doctoral Prize 2017. During her DPhil project, Oana has explored several different approaches to personalized search on the Social Semantic Web, in the form of preference-based ontological query answering. She considered single-user and group preferences, as well as different preference models, ranging from qualitative (unconditional and conditional) preferences to score-based preferences. Her work is based on the assumption that the preferences are given; the wide-spread applicability of her results in practice thus hinges on the existence of suitable preference mining technologies. 

During the EPSRC Doctoral Prize project, Oana will close this gap. She will explore and develop suitable techniques for mining preferences for personalized search on the Social Semantic Web. She especially wants to extract preferences from textual documents on the Web (such as blog posts, recommendations, and reviews) via deep-learning-based techniques for natural language processing, such as deep-learning-based sentiment analysis. She also wants to develop a small preference-mining demonstrator, which will be used for presentations to companies (towards potential collaborations with industry and the commercialization of the results of her dissertation) and for other outreach activities.

The EPSRC Doctoral Prize scheme is designed to help universities recruit the best PhD students, awarding them EPSRC support to increase the impact of their PhD, and to improve retention of these students in research careers.

Oana has finished her DPhil (PhD) project in the Information Systems group, under the supervision of Professor Thomas Lukasiewicz.