Context
Large-scale graphs are ubiquitous in many applications, ranging from transportation to social networks or data mining. Such large-scale graphs are naturally evolving over time, which makes particularly challenging data management tasks such as storage, querying, and mining. The LEG project aims to answer several research questions such as:
- How to construct a large evolving graph using temporal or spatio-temporal data?
- How to integrate large evolving graphs from heterogeneous sources and enrich them with semantic data?
- How to efficiently query a large evolving graph?
- How to store and analyze a large evolving graph in a distributed architecture?
The main application that motivates the LEG project is in the field of mobility analysis, although we aim at developing a generic approach.
The rest of the website contains details about the project members, meetings, publications, and software.
Members
- Engelbert MEPHU NGUIFO (Pr, project leader), Université Clermont Auvergne & CNRS LIMOS
- Sabeur ARIDHI (Mcf), Université de Lorraine & CNRS LORIA
- Dominique BARTH (Pr), Université de Versailles, DAVID lab
- Radu CIUCANU (Mcf), Université Clermont Auvergne & CNRS LIMOS
- Mamadou KANTE (Mcf Hdr), Université Clermont Auvergne & CNRS LIMOS
- Karine ZEITOUNI (Pr), Université de Versailles, DAVID lab
Meetings
- June 22, 2018: Meeting at LIMOS - Jounnée "Large Evolving Graphs" Participants: Engelbert Mephu Nguifo, Sabeur Aridhi, Karine Zeitouni, Jose Antonio de Macedo, Livia Almada, Lucas Peres, Remy Cazabet, Quentin Bramas, Wissem Inoubli, Yollande Eloise
- 10:00 - Engelbert Mephu Nguifo (LIMOS): Large Evolving Graphs
- 10:30 - Remy Cazabet (LIRIS, Univ Lyon): Tracking the evolution of communities in dynamic networks slides
- 11:15 - Quentin Bramas (ICube, Univ Strasbourg): L'évolution du graphe des transactions dans la cryptomonnaie IOTA slides
- 12:00 - Livia Almada (UFC Brazil and DAVID UVSQ Versailles): Location prediction on symbolic trajectories over road networks slides
- 12:30 - Break
- 14:00 - Jose Antonio Fernandes de Macedo (UFC Brazil): ODIN - Data Science applied to Police Intelligence slides
- 14:45 - Lucas Peres Gaspar (UFC Brazil): Fragment Extraction from RDFs Schemas using Keywords slides
- 15:15 - Break
- 15h45 - Yollande Eloise Mole Kamga (LIMOS): On Dynamic Graph Generation slides
- 16h15 - Wissem Inoubli (LIPAH, Tunisie): Scalable Interactive Dynamic Graph Clustering on Multicore CPUs slides
- 16:45 - Round table discussion around evolving graphs
- September 18, 2017: Meeting at the TDLSG workshop, which is organized by a subset of the LEG members and collocated with ECML/PKDD in Skopje, Macedonia. Photos taken at the workshop.
- July 25, 2017: Meeting in Paris, hosted at CNRS - Délégation Paris Michel-Ange
Participants: Engelbert Mephu Nguifo, Sabeur Aridhi, Dominique Barth, Radu Ciucanu, Mamadou Kante, Karine Zeitouni, Zoubida Kedad, Karim Alami
Program:- Talks
- Sabeur Aridhi: BLADYG: A Graph Processing Framework for Large Dynamic Graphs
- Radu Ciucanu: Synthetic Generation of Evolving Graphs
- Karine Zeitouni: Indexing In-Network Trajectory Flows
- System demonstration
- Karim Alami: EGG: A Framework for Generating Evolving RDF Graphs
- Round table discussion around evolving graphs
- June 27, 2017: Skype meeting
Program:
Publications
Demo paper at ISWC 2017:- Karim Alami, Radu Ciucanu and Engelbert Mephu Nguifo: EGG: A Framework for Generating Evolving RDF Graphs
- Karim Alami, Radu Ciucanu and Engelbert Mephu Nguifo: Synthetic Graph Generation from Finely-Tuned Temporal Constraints
- Ali Masri, Karine Zeitouni and Zoubida Kedad: Discovering Connections in Heterogeneous Transportation Datasets with Link++
- Abdelfattah Idri, Mariyem Oukarfi, Azedine Boulmakoul and Karine Zeitouni: Design and Implementation Issues of a Time-Dependent Shortest Path Algorithm for Multimodal Transportation Network
- Wissem Inoubli, Livia Almada, Ticiana Linhares, Gustavo Coutinho, Lucas Peres, Regis Pires Magalhaes, Jose Antonio F. de Macedo, Sabeur Aridhi and Engelbert Mephu Nguifo: A Distributed Framework for Large-Scale Time-Dependent Graph Analysis
Software
- EGG (Evolving Graph Generator) is a framework for generating evolving graphs based on finely-tuned temporal constraints given by the user, developed by the master student Karim Alami, under the supervision of Radu Ciucanu and Engelbert Mephu Nguifo.
