|NeSC Bibliographic Database|
Benchmarking ontology-based annotation tools for the Semantic Web
Appeared in: Proceedings of the UK e-Science All Hands Conference 2005 website: http://www.allhands.org.uk/2005/
Publisher: Engineering and Physical Sciences Research Council
Field of Science: e-Science
Abstract: Applications such as semantic annotation enable us to combine and associate existing ontologies in the biological field, and to perform an integral modelling of the disparate biological data sources. This then enables knowledge to be extracted from data repositories by means of agents, annotations and the semantic grid. This paper discusses and explores the main issues for evaluating ontology-based annotation tools, a key component in text mining applications for the Semantic Web. Text mining concerns the discovery of facts and structured information from large collections of unstructured documents. Semantic annotation and ontology-based information extraction technologies form the cornerstone of text mining applications for the Semantic Web. There has been a great deal of work in the last decade on evaluating traditional information extraction (IE) systems in terms of performance, but such methods are not sufficient for ontology-based systems, as we shall discuss. Furthermore, there are many other aspects of such tools which also need to be evaluated when comparing or establishing the usefulness of different systems, which go beyond the typical requirements for a traditional IE system, in terms of criteria such as usability, scalability, accessibility and interoperability. We outline the main requirements in terms of both usability and performance, proposing possible solutions and using examples of existing tools and resources.
Keywords: e-Science, AHM 2005
|Last Updated: 22 Jun 12 11:02|