PhD position in Semantic Web, Web of Data, Linked Data, Ontologies

PhD studentship proposal


Within the University de Lyon (France), the University Jean Monnet is 
searching for a full-time PhD student on the topic: A collaborative 
framework for ontology and instance data co-evolution and extraction for 
a joint PhD with the University of Bonn.


== Essential Facts ==

Connected Intelligence team: 
http://connected-intelligence.univ-st-etienne.fr/

Enterprise Information Systems team: http://eis.iai.uni-bonn.de/


The position is offered in the frame of a French-German doctoral college 
(SeReCo: Semantics, Reasoning and cooperation. 
https://sereco.univ-st-etienne.fr/) supported by the French-German 
University (DFH/UFA) and involving the research teams Connected 
Intelligence (Saint-Étienne, France) and Enterprise Information Systems 
(Bonn, Germany).

The topic of the PhD covers the domains: Semantic Web, Web of Data, 
Linked Data, ontologies, knowledge engineering, collaborative authoring, 
knowledge extraction, ontology evolution.

The PhD student will spend 3 years on the topic. He/She will have the 
chance to visit the University of Bonn (and the Fraunhofer IAIS). A 
double PhD degree with the University of Bonn is planned with this 
position (with at least 1 semester spent in Bonn).

The gross monthly salary is about 1684€ (up to 2000€ with tutoring) 
including social security etc.

Low tuition fees (about 400€/year), cheap housing possibilities, and 
addition scholarship for the mobility to Bonn (600€/month).


== Timeline ==

Application deadline: September 15th, 2016

PhD start: As soon as possible, depending on the administrative procedures.


== Abstract ==

Ontologies are at the heart of the Semantic Web, i.e. making data 
published on the Web comprehensible to intelligent added value services. 
Ontologies’ consensual design ensures its usefulness and wide acceptance 
by service developers. Collaboration of ontology engineers, domain 
experts from multiple disciplines, and end-users is required for 
defining them and populating them with instance data.

Collaborative ontology creation has been well studied, especially by our 
teams [1, 2, 3]. This PhD addresses the co-evolution of an ontology and 
its instance data. The bottleneck lies in identifying and assessing data 
originating from diverse sources (organisations, schemas, formats…) and 
managing co-evolution [4, 5]. Extensive automation of this process is 
desirable but still requires human collaboration and consensus, taking 
into account measures on data: quality, trust, data provenance. We will 
elaborate a generic approach covering all aspects of ontology 
co-evolution. It will integrate different means/techniques tailored to 
types of data sources (unstructured documents as well as specialized and 
generalist structured knowledge bases [6]), design a data model to 
characterize datasets, and implement methods for both automated and 
user-oriented collaborative processes.

We will experiment with the application of this foundational research to 
different fields, one of them being the field of independent music 
labels and artists. This field is well suited for our investigations 
because i) its structure and content are permanently evolving (new music 
genres, new albums, etc.), ii) its description has to take into account 
different people’s perspectives (experts, consumers), and iii) it is not 
yet well documented on the Web, which requires combining various data 
sources.


== Our references ==

[1] Maxime Lefrançois and Antoine Zimmermann, LinkedVocabularyEditor: 
une extension MediaWiki pour l'édition collaborative et la publication 
de vocabulaires liés, Demo at 26e Journées francophones d'Ingénierie des 
Connaissances, IC, June 2015, Rennes, France.

[2] Niklas Petersen, Lavdim Halilaj, Christoph Lange, and Sören Auer. 
VoCol: An Agile Methodology and Environment for Collaborative Vocabulary 
Development. Technical report, 2015. https://zenodo.org/record/15023/.

[3] Niklas Petersen, Irlán Grangel-González, Gökhan Coskun, Sören Auer, 
Marvin Frommhold, Sebastian Tramp, Maxime Lefrançois, Antoine 
Zimmermann: SCORVoc: Vocabulary-Based Information Integration and 
Exchange in Supply Networks. International Conference on Semantic 
Computing ICSC 2016: 132-139

[4] Riyadh Benammar, Alain Trémeau, Pierre Maret. An Approach for 
Ontology Population Based on Information Extraction Techniques - 
Application to Cultural Heritage. OTM Conferences 2015: 397-404.

[5] Sidra Faisal, Kemele M. Endris, Saeedeh Shekarpour, Sören Auer. 
Co-evolution of RDF Datasets. International Conference on Web 
Engineering (ICWE) 2016. http://arxiv.org/abs/1601.05270

[6] Antoine Zimmermann, Christophe Gravier, Julien Subercaze, Quentin 
Cruzille. Nell2RDF: Read the Web, and Turn it into RDF. KNOW@ LOD, 2-8. 
2013.


== Eligibility and application procedure ==

Candidates must hold a MSc degree in computer science or a related field 
and be able to combine both theoretical and practical aspects in their work.

Fluent English communication and a passion for developing modern 
software solutions are fundamental requirements. Command of French or 
German is a plus but not required. The candidates should have experience 
and commitment to work in team and on the forefront of research.

Candidates will need to submit by email a CV, university transcripts, a 
motivation letter including a short research plan targeting the above 
description, two letters of recommendation, and English writing samples 
(e.g., prior publication or master thesis excerpt).

The Université Jean Monnet is an equal opportunities employer. 
Preference will be given to suitably qualified women or persons with 
disabilities, all other considerations being equal.


== Contact ==

Pierre Maret and Antoine Zimmermann (Laboratoire Hubert Curien, 
Saint-Étienne, Connected Intelligence team: 
http://connected-intelligence.univ-st-etienne.fr/)


pierre.maret@univ-st-etienne.fr, antoine.zimmermann@emse.fr