Faites connaître cette offre !
Reference : UMR7030-HAIZAR-001
Workplace : VILLETANEUSE
Date of publication : Saturday, November 23, 2019
Type of Contract : FTC Scientist
Contract Period : 7 months
Expected date of employment : 1 February 2020
Proportion of work : Full time
Remuneration : Gross salary: 2650 €
Desired level of education : PhD
Experience required : 1 to 4 years
Ontologies are useful to overcome the heterogeneity issue and achieve the semantic interoperability between information sources. However, these ontologies are themselves heterogeneous, distributed and are described in different languages. Ontology alignment consists of a set of correspondences between semantic entities (concepts, roles), it bridges the semantic gap between ontologies and ensures semantic interoperability.
The labels that are used to express the entities of an ontology are in different languages. Most of the ontologies available on the web have an English lexical level. When publishing and linking an ontology on the web, we need to find correspondences with existing ontologies. These correspondences provide valuable knowledge in understanding and reusing ontologies.
In the context of PCU (Unified Knowledge Platform) project, LIPN proposed a product ontology built on the product database structure used in an e-commerce application. At this point, linking the produced ontology to ontologies in the same domain but in English is a core problem. It requires to propose robust algorithms that go far beyond translation as semantic labels are generally short and ambiguous.
The proposed algorithms will be tested on a standard benchmark (Multifarm), provided by the Ontology Alignment Evaluation Initiative (OAEI) campaign.
- state of the art update
- proposition and implementation of a multilingual ontologie alignment algorithm
- set on and analysis of experiments
- experience on building and alignement of ontologies and knowledge bases ;
- good skills on developing and semantic web technologies.
- Java / Python,
- Integrated Development Environment (such as ECLIPSE).
Knowledge Representation and Natural Language team focuses its research on natural language processing and knowledge engineering.
We talk about it on Twitter!