Research Topics: Pascale Sébillot has a background in natural language processing, extracting lexical and semantic knowledge from textual corpora using machine learning techniques, with applications to information retrieval and automatic speech recognition. Her current research interests are in NLP, multimedia collection modeling and structuring, combination of knowledge bases and multimedia content for data journalism.
Recent conference organization: co-chair of TALN 2018 program committee; member of the organizing committee of CORIA-TALN-RECITAL 2018
Recent and ongoing projects:
Brittany region grant (ARED)+AI for semantic data analytics ANR's program grant (AI4SDA) 2022-2025: detection of relations between entities from few or zero examples in press archives (scientific co-coordinator);
ANR Archival 2019-2023: development of multimedia archives and multimodal machine comprehension of language for new intelligent interfaces of scientific and cultural mediation;
JPI CH READ-IT 2018-2021: an ICT - HSS joint venture to explore EU's cultural heritage on reading by creating and curating a database of reading experiences;
Inria Project Lab iCODA 2017-2021: seamlessly integrating content, data and knowledge for interactive incremental analytics on heterogeneous data sources to help data journalism;
Bilateral contract with Ouest-France 2018-2021: combination of text mining and knowledge representation techniques to assist the extraction of valuable pieces of information from the journal's content so as to incrementally build a full-scale knowledge base (scientific co-coordinator);
Labex CominLabs LIMAH 2013-2018: a pluridisciplinary approach to linking media content for navigation and exploration (scientific co-coordinator).