- From: Julien Plu <julien.plu@redaction-developpez.com>
- Date: Mon, 16 Dec 2013 09:54:57 +0100
- To: public-lod@w3.org
- Cc: johannes.heinecke@orange.com
- Message-ID: <CAE94aYXU=8qfXBX_WOo7FjrnpH3-Kyhb5191mxje5HoX4uxV+w@mail.gmail.com>
Bonjour, Hello, Orange Labs (to Lannion in France) is currently offering a post-doc for a year. Applications are invited from France, EU or international young phd's (less than a year) in computer science or NLP. *Context* : A major challenge in computational linguistics is to generate a conceptual representation of documents by identifying the adequate “meaning” of terms. The varying concepts associated with the words constitute the conceptual space needed for further processing (extracting semantic relations, thematisation, …) *Subject* : The Natural Language Processing Team (CONTENT/FAST) of Orange Labs is currently working on information extraction in order to obtain a RDF representation. We base this work on existing knowledge bases (internal ones as well as community efforts like DBpedia or LinkedOpenData in general). The successful candidate will first extend our semantic data by merging it with public resources (linked opend data, linguistic linked open data like WordNet, BabelNet, DBpedia, Wiktionary etc). This increase should be quantitative by mainly adding named entities but also qualitative (using owl:sameAs) by targeting conceptual hierarchies (e.g. WordNet). In a second step, the candidate will implement access to this space by coupling it to our NLP tools (extraction of semantic links from a text). The overall treatment will be applied to corpora in order to study the impact of each resource on the result of an analysis. The result of the study should provide information to determine the actual value of these resources and how to use them together. Another, more fundamental, study will concern the concept granularity. The number and the nature of the various concepts associated to a term (word) is an open problem which depends on the data to be handled. *Requirements*: - a PhD in computer science or computational linguistics obtained during the last 12 months - knowledge in ontology alignment and the Semantic Web (technologies - (formats, ontologies, …) - knowledge in computational linguistics, syntactic and semantic analysis - Java, c++ programming languages, Linux - Working knowledge of French For details, please contact: Dr. Johannes Heinecke Phone: +33 2 96 05 21 77 Email : johannes.heinecke(at)orange.com Best Regards. Julien Plu.
Received on Monday, 16 December 2013 21:51:45 UTC