KEA - Knowledge Extraction Agent (»Add to Infobox)

Research Leader: Associate professor Hercules Dalianis Martin Hassel, Sumithra Vellupillai

Computer and Systems Sciences (Div)

During the last few years there has been an explosion of information on the Internet as well as a major push forward for digitalization of information within organizations. The digitalization has changed the working processes and also caused ergonomic and work environmental problems. Users of search tools need smarter tools to obtain information quickly but also to find hidden connections between different information sources. We at DSV/KTH-Stockholm University have followed this growth and developed a set of tools that both makes work easier for users in front of the screen and also finding hidden connection between texts. Automatic clustering, topic detection and named entity recognition are among the tools. Two organizations that are in need of these tools participate in our project, Stockholms läns landsting, SLL (Stockholm county council) and the business intelligence company Agent25. We are going to transfer this knowledge to these receivers and show them how to use the information, and simultaneously will we bring their problems and material back to us at DSV and solve new problems. SLL has several hundred thousand patient files that are going to be clustered according to content so we can find interesting relationships among them. Agent25 has over 40 million news articles that need to be tagged automatically with information such as personal name, product name, organization name and geographical locations. Period: 2006-11-01 - 2009-12-31 Keywords: Automatic text clustering, Business intelligence, Entity recognition, Medical informatics, Topic detection Project URL: http://dsv.su.se/en/research/ithealth/projects/kea/

SOURCE OF FUNDING (1/1)