PhD Thesis Final Defense to be held on 31 May 2018 at 14:00
(Image Source: https://thenextweb.com/socialmedia/2013/11/24/facebook-grandparents-need-next-gen-social-network/)
The examination is open to anyone who wishes to attend (Room 0.2.2, Decision Support Systems Laboratory, Old ECE Building)
The thematic area of the PhD dissertation at hand is the in-depth study of the way and the extent to which the social web, along with the phenomena that accompany it and the tools and technologies related to its exploitation and analysis can contribute to the policy cycle.
As far as the contribution of the dissertation is concerned, a study on whether tools and/ or techniques that have an applied in the scientific area of artificial intelligence can contribute to the exploitation of the very large volume of (structured and unstructured) information available in the social web, towards serving the needs of the different phases of the policy cycle. In particular, the dissertation focuses on:
• The potential contribution of argumentation techniques, so that the stakeholders can effectively and efficiently exploit the "wisdom of the crowd" for the needs of the policy cycle.
• The potential contribution of causal models, mainly focusing on fuzzy cognitive maps, in order to take advantage of existing information and existing data towards both reasoning on the outcomes of past policies, as well as being able to predict the potential impact of future initiatives.
Towards testing and validating the dissertation’s proposals/ hypotheses regarding argumentation techniques, five (5) carefully selected unstructured dialogues from the Greek social web were analyzed, leading to valuable conclusions on both the techniques to be applied so that the analyst can effectively and efficiently exploit unstructured dialogues, as well as on the way unstructured dialogue takes place in the Greek social web in general.
Towards testing and validating the proposals/ hypotheses regarding causal models, three (3) alternative scenarios on the Greek economy were designed and run, leading to valuable conclusions in relation to the suitability of fuzzy cognitive maps (as well as causal models in general) in the policy cycle.
In addition, the dissertation aims to contribute to the cornerstone of not just policies, but also of any initiative towards creating added value: requirements. Therefore, a framework consisting of a set of ten (10) requirements that can guide conception, design and implementation of policies is presented.
Supervisor: John Psarras, Professor
PhD student: Kokkinakos Panagiotis