2014 IntegrationofLargeScaleKnowledg

Subject Headings: Large Knowledge Base, KB Mapping Task.

Notes

Over the recent past, information extraction (IE) systems such as Nell and ReVerb have attained much success in creating large knowledge resources with minimal supervision. But, these resources in general, lack schema information and contain facts with high degree of ambiguity which are often difficult to interpret. Whereas, Wikipedia-based IE projects like DBpedia and Yago are structured, have disambiguated facts with unique identifiers and maintain a well-defined schema. In this work, we propose a probabilistic method to integrate these two types of IE projects where the structured knowledge bases benefit from the wide coverage of the semi-supervised IE projects and the latter benefits from the schema information of the former.

Research in information extraction (IE) systems has experienced a strong momentum in recent years. While Wikipedia based information extraction projects such as DBpedia [1, 16] and Yago [23] have been in development for several years, systems such as Nell [5] and ReVerb [9] that work on very large and unstructured text corpora have more recently achieved impressive results. The developers of the latter systems have coined the term open information extraction (OIE) [2] to describe information extraction systems that are not constrained by the boundaries of encyclopedic knowledge and the corresponding _xed schemata that are, for instance, used by Yago and DBpedia. The data maintained by OIE systems is important for analyzing, reasoning about, and discovering novel facts on the web and has the potential to result in a new generation of web search engines [8]. In this context, there are latent advantages in integrating these two IE systems in producing better knowledge repositories which can be harvested to further enable state-of-the-art performance on a wide range of NLP applications. This is especially promising, since Nell and ReVerb typically achieve a very large coverage, but still lack a full-edged, clean ontological structure which, on the other hand, could be provided by large-scale ontologies like DBpedia or Yago.

;

	Author	volume	Date Value	title	type	journal	titleUrl	doi	note	year
2014 IntegrationofLargeScaleKnowledg	Arnab Kumar Dutta			Integration of Large Scale Knowledge Bases Using Probabilistic Graphical Models				10.1145/2556195.2556202		2014