While at Yahoo! Research I was involved in various projects on analyzing, extending and utilizing tagging of Flickr photos. We analyzed what type of tags users applied to their photos and devised a method for classifying those tags into semantic categories. We developed two prototypes, one for tag recommendation and and one for tag based browsing of photos. We started our tag analysis by mapping tags to WordNet broad categories and found out that while users annotations are sparse, they…
One of my most fruitful project while working at Yahoo! Research was a collaboration with the Yahoo! Image Search team. We built system for extracting and ranking points of interest for towns, countries and regions. The points of interest were extracted from Wikipedia, Yahoo! GeoPlanet and Yahoo! Travel. The ranking was based on co-occurrence statistics from Flickr photos and Yahoo! Image Search query logs. The resulting data package was generated by the research team and applied by Yahoo! Image Search…
Simon Overell, Börkur Sigurbjörnsson, and Roelof van Zwol.
Classifying Tags using Open Content Resources.
In: Proceedings of the Second ACM International Conference on Web Search and Data Mining (WSDM 2009).
The main research topic during my PhD research was XML Element Retrieval . I took a corpus of scientific articles in XML format and built a search engine on top of it to run experiments. For a given user query, the task was to go below the document level and return the XML elements that best matched the users’ information need. Most of the initial research effort went into evaluating different search concepts against the INEX XML retrieval evaluation collection….
Börkur Sigurbjörnsson, Jaap Kamps, and Maarten de Rijke.
Focused Access to Wikipedia.
In: 6th Dutch-Belgian Information Retrieval Workshop (DIR 2006).
Wikiii was a search engine interface for giving focused access to Wikipedia. It was powered by the Lucene based XML element retrieval engine I developed as part of my PhD research. For a given query the search engine gave direct access to the most relevant sections of the relevant Wikipedia articles.