WEBINAR DOING@MADICS

L’atelier DOING invite les chercheurs de la communauté MADICS à participer de ses webinars.

WEBINAR 2 – 17 JUIN 2020

Twitter : https://twitter.com/NetworkDoing

Canal Slack #doing-webinar-2 :  https://join.slack.com/t/doing-madics/shared_invite/zt-eg5yg240-pqu~Q2igaDOXs1OpmzOOHg

PROGRAMME:

  • 10:00 – 11 :00  : Keynote: Natural Language Processing in the Health Domain, Dr. Vasiliki Foufi, University of Geneve
  • 11:00 – 11:25  :  On Anonymizing the Provenance of Collection-Based Workflows, Khalid Belhajjame, PSL, Université Paris-Dauphine, LAMSADE
  • 11:25 – 11:45 : Exploring COVID-19 documents with the S-COVID engine, Mehrdad Farokhnejad, Université de Grenoble, LIG

KEY NOTE

Abstract: Digitalization is transforming all aspects of everyday life, it is also deeply transforming medicine. Nowadays, there is an enormous and rapidly growing amount of health-related free-text data coming from various sources: electronic health records, scientific literature, social networks. Free-text data (also called unstructured data) contain important health information but their processing constitutes a great challenge for data scientists and physicians mostly due to their heterogeneous and “unstructured”characteristics. Another important aspect of health-related data is their use for research purposes while ensuring data security and privacy and preserving data integrity. In this presentation, these challenges as well as techniques for processing free-text health-related data will be discussed.

Vasiliki Foufi is Research and Teaching Fellow in the Department of Radiology and Medical Informatics in the University of Geneva. She holds a Phd in Computational Linguistics and a Master in Applications of  Technology in Language Sciences and Communication. Her research interests lie in the field of Natural Language Processing (NLP) and its applications in different domains such as the health domain. Since 2003, she has been conducting research in information extraction, named-entity recognition, corpus annotation and parsing using NLP tools. She has worked on the development of linguistic resources such as electronic dictionaries, finite state automata and annotated corpora for French, English, Greek, German and Italian.

WEBINAR 1 – 5 JUIN 2020

Twitter : https://twitter.com/NetworkDoing

Nous vous demandons de vous inscrire AVANT le 4 Juin MIDI, via l’evento suivant en utilisant votre adresse e-mail institutionnelle : https://evento.renater.fr/survey/webinar-doing-1-46wq0pan

Canal Slack #doing-webinar-1 : 
https://join.slack.com/t/doing-madics/shared_invite/zt-eg5yg240-pqu~Q2igaDOXs1OpmzOOHg

PROGRAMME:

  • 10:00 – 11 :00  : Keynote: Managing data quality in the age of big data. Prof. Salima Benbernou, Université Paris Descartes, LIPADE.
  • 11:00 – 11:25  : Alignement de bases de données pour l’extraction d’informations concernant les sols pollués. Chuanming Dong, LASTIG, Univ Gustave Eiffel, ENSG, IGN (SLIDES)
  • 11:25 – 11:45 : DOING@DEFT : cascade de CRF pour l’annotation d’entités cliniques imbriquées. Andreane Roques, Université d’Orléans, LIFO (SLIDES)

KEY NOTE

Abstract: Managing data quality is  highly relevant to big data and other valuable new data assets from Web applications, social media, the digital supply chain, SaaS apps, and the Internet of Things. Today, the organisations  are harnessing multi-sources data to rise the benefits of their business. Therefore, the quality in big data is a cornerstone to operational, transactional processes and to the reliability of   business analytics for decision making. Unless an organization handles the quality, it may fail to deliver the kind of trusted analytics, operational reporting, self-service functionality, business monitoring, and governance that are expected of all data assets. In the talk  we will discuss different dimension of data quality: how to query such big amount of data efficiently and  harvest relevant and complete query results through data fusion process? how to discover the inconsistencies  because big data platforms invariably end up with the same data loaded multiple times with different values? and  how to  reason with the uncertainty often associated to data because the  fusion of conflicting data sources, measurement inaccuracy, sampling discrepancy, outdated data sources.

Salima Benbernou is full Professor in computer science at Université de Paris (Université de Recherche Intensive). Before that, she was an Associate Professeur at Université de Lyon. She is leading data intensive and kNowledge oriented system group (diNo) at LIPADE Lab. Her research interests include large scale data management, data quality, data privacy and data fairness.The interdisciplinary plays a key role in her research. Her work is supported by European Commission, National and Industrial agencies.