The objective of my PhD project is to develop and evaluate word cloud generation algorithms to enhance relevance, diversity and other aspects of these visual IR interfaces. These algorithms have a wide scope of applications in real word scenarios when there is a need to browse and explore a large number of documents.
The first part of the project is focus on tag cloud generation from folksonomy data generated within social tagging systems. Standard word cloud generation approaches fail to provide relevant tags and therefore prevent user retrieval of relevant content. The latter part of the project is focused on social media services such as Twitter. We define several extensions of standard word cloud generation for micro blogging data which help to tackle a problem of information overload when browsing a large number of tweets.

During this PhD project, I have been involved in the research collaboration with Danish tax office called Skat. I believe that prior knowledge from AR would be beneficial for more effective progress during the project. In particular, it was not clear to me how to define an agreement between Skat organization and my PhD research such that it will be beneficial from both sides. Further, was the collaborative involvement or iterative process the ideal project choices?

After my PhD, I am planning to go to industry where I would like to try to reduce the gap between an academic research and the practices exploited in the industry. I believe that AR could provide me guidelines how to reduce this knowledge gap.