Named Entity Recognition using FOX

Unstructured data still makes up an important portion of the Web. One key task towards transforming this unstructured data into structured data is named entity recognition. We demo FOX, the Federated knOwledge eXtraction framework, a highly accurate open-source framework that implements RESTful web services for named entity recognition. Our framework achieves a higher F-measure than state-of-the-art named entity recognition frameworks by combining the results of several approaches through ensemble learning. Moreover, it disambiguates and links named entities against DBpedia by relying on the AGDISTIS framework. As a result, FOX provides users with accurately disambiguated and linked named entities in several RDF serialization formats. We demonstrate the different interfaces implemented by FOX within use cases pertaining to extracting entities from news texts.