KYOTO Logo
Knowledge Yielding Ontologies
for Transition-based Organization
  • Increase font size
  • Default font size
  • Decrease font size

D5.4 Fact Miners - revised

Downlad PDF

This document describes the Fact Mining process in KYOTO. The knowledge contained in the domain texts is automatically extracted by means of the new Information Extraction technology developed in KYOTO. This technology can be used to extract from the domain texts both conceptual and factual knowledge. On the one hand, conceptual knowledge is expressed through properties and relationships among domain terms (described in deliverable D5.3 Concept Mining revised). On the other hand, factual knowledge is generally expressed through properties and relationships, either eventual or not, among specic entities occurring in the domain text.

For such a goal, KYOTO developped Tybots or Term Yielding Robots for extracting conceptual knowledge, and Kybots or Knowledge Yielding Robots for extracting factual knowledge. Kybots incorporate technology for the extraction of relationships, relative to the general or domain concepts already captured by the Tybots. That is, the extraction of factual knowledge will be carried out by the Kybot server by processing Kybot proles on the linguistically enriched documents.

First, this document provides a comprehensive survey on Information Extraction and the cur- rent Linguistic Annotation standards. Then, it concentrates on the design of the Kybot architecture and presents the main capabilities of the Kybot mining module, the tool that we use for fact extraction. Finally, it provides the results of the mining module evaluation and several important issues regarding implementation details.

 

ICT-211423 - 2008 © Kyoto Consortium