作者: Maria Vasardani , Stephan Winter , Junchul Kim
DOI:
关键词:
摘要: This paper proposes a novel approach of harvesting large corpora from the Web for generating place graphs. The consists three main phases: an ecient strategy relevant web pages that include descriptions related to particular environment, extracting triplets descriptions, and graphs these triplets. also discusses characteristics generated graphs, identies further challenges given well- known exibility natural language.