Digital Humanities Laboratory (DHLab)

The DHLab: A Data Science Center supporting Digital Humanities

To maintain the emphasis on transdisciplinarity, which is the main focus of all actions carried out through the IDEX UCAJEDI project, the Center of Modeling, Simulation, and Interactions (MSI) has created its second thematic expertise center after the MDLab: The Digital Humanities Lab (DHLab).  

The DHLab aims to strengthen ties between data scientists and researchers in social sciences and the humanities at Université Côte d’Azur.

Objectives

The activities of the DHLab fall under the Data Science core program of the UCAJEDI project. The DHLab is currently running two "proof of concept" projects (“DeepText” and “Regional expertise alignment and firm performance”), both addressing text mining issues. Once this pilot phase is completed in 2020, the DHLab will support research and development within Université Côte d'Azur on other modeling and data processing issues related to social sciences and humanities. It will provide valuable assistance in modeling, classifying and visualizing heterogeneous data.

Are you a researcher in the social sciences and humanities who needs to process massive data? The DHLab will assist you in writing a protocol or a research project in order to submit it to a call for proposals assessed by DHLab’s strategic committee.

Under the responsibility of this strategic committee, the DHLab will also back projects already scientifically approved by Université Côte d'Azur and potentially involving private partners. In addition, the DHLab will offer training programs focused on the new skills required to deal with data science in the social sciences and humanities. If you wish to apply for national or European funding (ANR, ERC,...) and your project involves recruiting a data scientist, you can contact the DHLab executive director who will help you to write the section of the project that describes the engineer's tasks. Once the project has been accepted, the MSI will also help you to recruit the engineer and let him/her benefit from MSI start-up funds.

Training

The DHLab is currently contributing to the deployment of a cycle of short training courses, organized as seminars and workshops on data science, artificial intelligence, and data integration applied to several domains, held at the MSI site in Sophia Antipolis.

Find out more about the MSI's short training courses and subscribe to the MSI’s newsletter!

Team

DHLab strategic committee chaired by Arnaud Zucker, Prof. of Greek language and literature, and consisting of Université Côte d’Azur representatives, is in charge of the decisions relating to DHLab’s activities, in consultation with the UCAJEDI project governance.
The following DHLab staff is working under the hierarchical authority of the Director of the MSI, Prof. Stéphane Descombes:

  • Dr. Marco Corneli, DHLab operational manager, ensures the proper functioning of the center and is involved in the Deep Text project, carried out at the Bases, Corpus, Langage (BCL) laboratory of Université Cote d’Azur.
  • Jean-Noël Mattei, data scientist, is working on the project "Regional expertise alignment and firm performance” at Skema Business School (Sophia Antipolis campus).
Areas of Research

The DHLab's areas of research are as follows:

  • Pre-processing algorithms and unstructured data cleansing
  • Classification, processing and modeling of heterogeneous data such as texts and interaction data (graphs)
  • Deep learning, convolutional or recurrent networks.

Projects can benefit from the databases of the following entities:

  1. Université Côte d'Azur laboratories;
  2. INSEE;
  3. European Patent Office;
  4. Microsoft Academic.

Publications

Preprints

L. Vanni, M. Corneli, D. Mayaffre, F. Precioso: From text saliency to linguistic objects: learning linguistic interpretable markers with a multichannel convolutional architecture (2020).

Journal Papers

M. Corneli, C. Bouveyron, P. Latouche: Co-Clustering of Ordinal Data via Latent Continuous Random Variables and Not Missing at Random Entries, in “Journal of Computational and Graphical Statistics,” March 2020.

Book Chapters

L. Vanni, M. Corneli, D. Longree, D. Mayaffre, F. Precioso: Text Analytics, Chapter 4: Key passages: from statistics to deep learning, November 2020.
 

Contacts

Dr. Marco Corneli:  Marco.CORNELI[at]univ-cotedazur.fr