FRED and Tipalo: from natural language to RDF/OWL
We are pleased to announce the release of two new tools: FRED and Tipalo. Online demonstrators are available from the STLab tools page that collects STLab software releases. We invite you to play with them, and provide your welcome feedback:
A tool for automatically producing RDF/OWL ontologies and linked data from natural language sentences, currently limited to English.
FRED is based on C&C and Boxer, a NLP tool that transforms natural language text into a logical form compliant to Discourse Representation Theory. We process Boxer output and apply a set of heuristics and semantic transformations in order to obtain RDF designed for the Semantic Web. In this process, we emphasize the relation to linguistic frames, supporting FrameNet and VerbNet vocabularies, and to ontology design patterns.
In order to further improve interlinking of FRED results with LOD, a number of features are under testing (they are already available in the Tipalo tool), including Named Entity Resolution (based on Apache Stanbol) and Word Sense Disambiguation (based on UKB). FRED results are available as n-triples or graphs. A paper on FRED will be presented at next EKAW2012.
A tool that automatically assigns types to Wikipedia entities in a LOD-intensive graph. Given a Wikipedia page URI, the tool returns a RDF graph composed of rdf:type, rdfs:subClassOf, owl:sameAs, and owl:equivalentTo statements providing typing information (organized into class taxonomies) about the entity referred by the Wikipedia page. Currently, entity types are derived from the text, and aligned to the DBpedia Ontology, WordNet 3.0 in RDF, DUL, and DolceZero. The tool relies on FRED (including NER and WSD), hence it has unlimited domain coverage (i.e., it is independent from the completeness of specific ontologies). Results are available as RDF, HTML (with LODE), and graphs. A paper describing Tipalo is currently under review.
Each tool is described in more detail in dedicated documentation.
Feedback welcome to email@example.com
The STLab team
This work has been partly supported by EU IKS project and developed in collaboration with the Computer Science department of the University of Bologna.