Position : Professor on Computer Science (position #20)

Teaching

The recruited person will be expected to teach in all fields of the Computer Science department of the UFR Sciences d’Orsay, at the Bachelor’s and Master’s levels (classical and apprenticeship). In particular, she/he should be able to teach data science and/or artificial intelligence at the bachelor’s and master’s level. Target courses include (but are not limited to): NLP and dialogue systems, information retrieval and hybrid reasoning (RAG, neurosymbolic approaches), multimodal learning and constraint learning, graph structure and analysis, knowledge acquisition and representation, large scale data management and querying, multi-agent and adaptative systems, algorithms and programming (neural architectures, deep learning methods, Python), generative models, applied statistics, advanced optimisation and automatised machine learning (AutoML), as well as ethical and responsible AIArtificial Intelligence (fairness, bias, environmental and societal impacts).
The person hired may also be required to teach part of the courses in English, particularly in the context of the Master degrees in computer science or bioinformatics, and to take part in the supervision of multidisciplinary projects (bioinformatics, linguistics, sociology, health, etc.).
A strong commitment for teaching in the first years of undergraduate studies is required. The Computer Science department training offer is available at:
https://ecole-universitaire-paris-saclay.fr/formation/licence/informatique #parcours for Bachelor level.
Teaching is one of the university’s core missions. More than ever, the quality of training provided and the quality of student learning are at the heart of Université Paris Saclay’s concerns. As such, the teaching profile of this position includes teaching courses, tutorials (“TDTravaux Dirigés (Tutorials)”) and practical work (“TPTravaux Pratiques (Practical work)”), as well as developing and updating course content in line with technological developments and student needs, supervising student projects (tutored projects, final projects) and internships in companies or laboratories, participate in setting up new training courses and upgrading existing programs, and use innovative teaching methods adapted to students’ needs (e-learning, inverted pedagogy, use of platforms, etc.).
The person recruited will also be involved in collective responsibilities related to teaching, in particular as future head of the Master program in AIArtificial Intelligence.

Research activities

The Laboratoire Interdisciplinaire des Sciences du Numérique (LISN – CNRS Joint Research Unit 9015, Inria, CentraleSupélec) was established in 2021 by bringing together 16 research teams from LIMSILaboratoire d'Informatique pour la Mécanique et les Sciences de l'Ingénieur, créé en 1972 et dont les équipes ont rejoint celles du LRI en 2021 pour fonder le LISN. and LRI (organized into five departments), along with their research support services, for a total of more than 300 people.
LISN is the Université Paris-Saclay laboratory with the strongest expertise in artificial intelligence: it includes over 170 members working in this field, more than 60 of whom are permanent staff, mainly in the Data Science (SDData Sciences) and Language Science and Technology (STL) departments.

LISN benefits from a supportive ecosystem (DataIA cluster, DeMythif.AIArtificial Intelligence COFUND PhD program, and highly selective, internationally oriented master’s programs in AIArtificial Intelligence). New hires can count on financial support from DataIA to launch their projects at Paris-Saclay, as well as support from the university’s Springboard program (potentially up to 120k€).

Building on this momentum, LISN is recruiting Professors and Associate Professors in AIArtificial Intelligence as part of a multi-year plan, with teaching duties at the Faculty of Science (UFR Sciences) of Université Paris-Saclay. Successful candidates will join the A&O, BIOINFO, orLaHDAK teams in the SDData Sciences department, or the LIPS, M3, or SEME teams in the STL department.

The candidate will get involved in the Paris-Saclay AIArtificial Intelligence ecosystem — for example, in the governance of DataIA, in its international outreach, as well as in LISN’s scientific life, potentially including the creation of a new AIArtificial Intelligence research theme and taking on responsibilities at the integrated department level. International standing will be appreciated. Candidates are invited to contact the heads of the departments and teams they wish to join (details here: https://www.lisn.upsaclay.fr/recherche/lia-au-lisn/ ).

Overview of the STL and SDData Sciences departments and their priority themes

The Language Science and Technology department (STL) seeks to strengthen activities aimed at developing AIArtificial Intelligence models and methods that contribute to discovering fundamental properties of language and to the effective analysis of written, spoken, or signed utterances. The STL department brings together three internationally recognized teams (LIPSLangue Interaction Parole et Signes, M3, SEME) that develop both statistical and symbolic AIArtificial Intelligence methods, with a multidisciplinary perspective combining computer science, signal processing, and linguistics. The department investigates fundamental questions regarding linguistic systems, leveraging large corpora collected, annotated, and enriched in unsupervised or semi-supervised ways. In doing so, we develop major language-processing applications (speech recognition, information retrieval, conversational agents, …) that carry increasingly significant societal and ethical implications. The department addresses issues of accessing meaning in linguistic productions, with objectives of analysis, understanding, modeling, and generation. We apply our research to written, spoken, and signed modalities and across a variety of registers and specialized domains, such as the biomedical field.

STL

Candidates for STL should present a research project in Natural Language Processing (NLP) and an integration plan aligned with the department’s priority themes around the study of large language models (LLMs), for example:

• Methods to account for the environmental impact of producing and using LLMs, with a view to generalizing sustainable research practices;
• Methods for adapting generative AIArtificial Intelligence to resource-scarce contexts (low-resource languages, few-shot learning), and for integrating knowledge to develop specialized and reliable LLMs: RAG, exploitation of terminological resources, information extraction;
• Study of the intrinsic properties of large language models (LLMs), particularly regarding bias and the preservation of privacy related to training or fine-tuning data;

Study of the ethical and social aspects of digital technology, in particular the impact of digital tools on science: epistemological issues, automatic evaluation of articles, fact-checking.

SDData Sciences department

The Data Science department (SDData Sciences) brings together four teams with recognized and complementary expertise (A&O, Bioinfo, LaHDAK, ROCS). They cover the entire pipeline of data and knowledge exploration—from modeling through collection, management, analysis, structuring, and exploitation via machine-learning methods. This complementarity fosters synergies around themes related to data, knowledge, machine learning, and optimization, with notable applications in simulation, bioinformatics, and the web.
Beyond the theoretical foundations and methodologies of data science, the department asserts a strong applicative dimension by contributing to major societal challenges such as health, ecological transition, life sciences, and digital infrastructures. This application focus goes hand-in-hand with ethical vigilance in research choices, with a commitment to developing frugal, explainable, and inclusive methods attentive to biases and their impacts. The department also stands out for its affirmed commitment to open and
reproducible research, fostering transparency, knowledge-sharing, and broad dissemination of scientific results.

AIArtificial Intelligence themes in the department include:

• Machine learning: “Good AIArtificial Intelligence” (frugality in AutoML, absence of bias/fairness, explainability, causality, etc.) and “AIArtificial Intelligence for Good” (for a sustainable society); interactions between statistical physics and learning (notably to understand training dynamics); incorporation of knowledge into learning (notably for numerical simulations and time series) (A&O team);

• Hybrid AIArtificial Intelligence, constraints, and massive data: developing methods that combine machine learning, symbolic reasoning (rules, constraints, ontologies, graphs), declarative and interactive data mining, as well as the management of massive and heterogeneous data. The goal is to design explainable, reliable, and efficient intelligent systems capable of acquiring and integrating knowledge, reasoning, and solving complex problems (LaHDAK);
• Learning for biology, particularly in evolution, genomics, and health (BioInfo).

This list is not exhaustive; new topics are welcome. More details on the AIArtificial Intelligence research themes of the departments and teams, as well as contact persons, are available on this page:

https://www.lisn.upsaclay.fr/recherche/lia-au-lisn/.