Service-oriented information extraction
Katja Seidler, Alexander Schil · 2011
The growth of unstructured data being available in the Internet, private and enterprise file shares demands flexible and generic solutions to extract the included information. Current information extraction solutions focus on domain specific approaches manually built together and adapted for a preassigned task. Moreover they suffer from uncertainty and inaccurate results, especially when the domain slightly changes. In this paper, we propose a new service-oriented information extraction approach to overcome the present disadvantages. We outline the major topics involved in the development of such a system and point to research questions we want to address in the doctoral work.