| Abstract
|
There is a lot of 'semantic information' o … There is a lot of 'semantic information' on the Web, the vast majority of which is encoded as human language text. This is especially true for content found on the social web, consisting of blogs, Wikis, forums, and many other social media systems. One way to accelerate the realization of the Semantic Web's vision of a web of machine understandable data is to extract semantic information from this text and publish it in structured or semi-structured forms (e.g., RDF) using appropriate ontologies. I will describe a range of current tools and efforts to do this, such as DBpedia, Freebase and Open Calais, as well as our own work Wikitology -- a hybrid knowledge base constructed from Wikipedia and other knowledge sources. rom Wikipedia and other knowledge sources.
|
