A5082193370- Semantic Web and Ontologies
- Data Quality and Management
- Biomedical Text Mining and Ontologies
- Advanced Database Systems and Queries
- Service-Oriented Architecture and Web Services
- Natural Language Processing Techniques
- Rough Sets and Fuzzy Logic
- Data Mining Algorithms and Applications
- Topic Modeling
- Data Management and Algorithms
- Computational Drug Discovery Methods
- Advanced Graph Neural Networks
- Web Data Mining and Analysis
- Personal Information Management and User Behavior
- Bayesian Modeling and Causal Inference
- Access Control and Trust
- Context-Aware Activity Recognition Systems
- Cognitive Computing and Networks
- IoT and Edge/Fog Computing
- Artificial Intelligence in Education
- Complex Network Analysis Techniques
- Bioinformatics and Genomic Networks
- Microbial Metabolic Engineering and Bioproduction
- Scheduling and Optimization Algorithms
- Speech and dialogue systems
Université Paris-Saclay
2012-2024
Centre National de la Recherche Scientifique
2012-2024
Laboratoire de Recherche en Informatique
2013-2022
Laboratoire Interdisciplinaire des Sciences du Numérique
2020-2022
Université Paris-Sud
2005-2019
Badji Moktar-Annaba University
2015
Université Paris Cité
2014
Institut national de recherche en informatique et en automatique
2005-2012
Inria Saclay - Île de France
2010-2011
Laboratoire d'Informatique, de Robotique et de Microélectronique de Montpellier
2008
Most of the Linked Data applications currently rely on use owl: sameAs for linking ontology instances. However, several studies have noticed multiple misuses this identity link. These misuses, which are mainly caused by lack other well-defined alternatives, can lead to erroneous statements or inconsistencies. We propose in paper a new contextual link: identiConTo that could serve as replacement identical instances specified context. To detect these links, we defined an algorithm named DECIDE...
Many approaches have been defined to link data items automatically. Nevertheless, when are numerous and the schema is unknown, most of these too time-consuming. We propose an approach where classification rules learnt thanks a training set made linked data. These can then be applied in order classify reduce linking space i. e item pairs that compared. First experiments conducted on RDF sets describing electronic products.
In recent years, keen interest towards Knowledge Graphs has increased in both academia and the industry which led to creation of various datasets development different research topics. this paper, we present an approach that discovers differential causal rules Graphs. Such express for two class instances, a treatment leads outcomes. Discovering is often key experiments, independently their domain. The proposed based on semantic matching relying community detection strata can be defined as...
Thanks to the initiative of Linked Open Data, RDF datasets that are published on Web more and numerous. One active research field currently concerns problem finding links between entities. We focus in this paper ontology-based data linking approaches which use rules based available schemas (or ontologies). This kind systems assume have beforehand a set mappings ontology elements. However, could be incomplete. propose approach called N2R-Part. It is computation similarity scores by exploiting...
This foreword introduces a summary of themes and papers the Web2Touch (W2T) 2020-21 Track at 29th IEEE WETICE Conference held as virtual Conference, in October 2020. W2T includes six full four short papers. They all address relevant issues field information sharing for collaboration, including, big data analytics, knowledge engineering, linked open data, applications smart Web technologies, care. The portfolio hot research semantics, technologies (e.g., IoT, sensors, devices tele-monitoring,...
In the absence of a central naming authority on Semantic Web, it is common for different data sets to refer same thing by names. Whenever multiple names are used denote thing, owl:sameAs statements needed in order link and foster reuse. Studies that date back as far 2009, observed property sometimes incorrectly. our previous work, we presented an identity graph containing over 500 million explicit 35 billion implied statements, scalable approach automatically calculating error degree each...