Acquiring Geographical Entities and their Relations from Texts
Cao Cun · 2007
There are various sorts of knowledge contained in text, and knowledge acquisition from text (KAT) has become one of the important research problems in knowledge engineering. How to acquire Geographical entities and their relations from large-scale Webpages is the research topic in this paper. Firstly, grammatical patterns are acquired automatically and semi-automatically to construct the Extraction Pattern Base. Then quasi-concept are extracted and verified locally with those resulting sentences generated from extraction patterns. At last, a Geographical Graph is defined based on graph theory, with the geographical rules to verify geographical entities and their relations. As an important part of the Conceptual Space managed by the Concept Graph, geographical entities and their relations are not only the important part of the Knowledge Base, but also provide support for other domain or domain independent knowledge.