You are here
STI International PhD Award Winners
Periodically, STI International runs a PhD student of the year award. This competition is open to all promising junior researchers within the STI International network who defended their thesis within a particular calendar year. The evaluation is carried out by senior academics who usually include one or more STI Fellows. The evaluation criteria are based on the scientific and general impact of the PhD work as well as any additional achievements the junior researcher may have.
Winners are given a cash award and also present their work at an STI International event.
Tom Heath
Winning Year of Award:
2009
Title of PhD thesis:
Information-seeking on the Web with Trusted Social Networks – from Theory to Systems
Abstract:
This research investigates how synergies between the Web and social networks can enhance the process of obtaining relevant and trustworthy information. A review of literature on personalised search, social search, recommender systems, social networks and trust propagation reveals limitations of existing technology in areas such as relevance, collaboration, task-adaptivity and trust.
In response to these limitations I present a Web-based approach to information-seeking using social networks. This approach takes a source-centric perspective on the information-seeking process, aiming to identify trustworthy sources of relevant information from within the user's social network.
An empirical study of source-selection decisions in information- and recommendation- seeking identified five factors that influence the choice of source, and its perceived trustworthiness. The priority given to each of these factors was found to vary according to the criticality and subjectivity of the task.
A series of algorithms have been developed that operationalise three of these factors (expertise, experience, affinity) and generate from various data sources a number of trust metrics for use in social network-based information seeking. The most significant of these data sources is Revyu.com, a reviewing and rating Web site implemented as part of this research, that takes input from regular users and makes it available on the Semantic Web for easy re-use by the implemented algorithms.
Output of the algorithms is used in Hoonoh.com, a Semantic Web-based system that has been developed to support users in identifying relevant and trustworthy information 2 sources within their social networks. Evaluation of this system's ability to predict source selections showed more promising results for the experience factor than for expertise or affinity. This may be attributed to the greater demands these two factors place in terms of input data. Limitations of the work and opportunities for future research are discussed.
Affiliation:
Image:
Website:
Chris Bizer
Winning Year of Award:
2008
Title of PhD thesis:
Quality-Driven Information Filtering in the Context of Web-Based Information Systems
Abstract:
Web-based information systems provide access to information originating from multiple information providers inside the company, from partner organizations, and from the public Web. The quality of provided information may vary as information providers have different levels of knowledge, different views of the world, and different intentions. Thus before information is used to accomplish a specific task, its quality should be assessed according to task-specific criteria. The goal of the thesis is to develop and evaluate a quality-driven information filtering framework which supports information consumers in their decision whether to accept or reject information. The main objective of the framework is to allow information consumers to apply a wide range of different subjective filtering policies. In order to facilitate the information consumers understanding of filtering decisions, the framework can generate explanations why information satisfies a specific policy. The framework is integrated into a web browser and is applied within a financial information integration scenario.
Affiliation:
Image:
Members|Partners
_itok-BIq--_jT.png "Karlsruhe Institute of Technology (KIT)")
