these : Interactive Explainability of Machine learning applied to language tasks



  • philippe muller [email protected]


  • Interactive Explainability of Machine learning applied to language tasks


  • **Project context**: The thesis takes place within the Descartes project (https://www.cnrsatcreate.cnrs.fr/descartes/), a large France-Singapore collaboration project on applying AI to urban systems.


  • The project will generate a lot of data about artifical systems deployed in the wild, part of which will be expressed as textual data (expert reports, user reactions, news coverage, social media conversations).


  • Natural language processing (NLP) models can help access that voluminous information, but there is an important need from operators, policy makers and public institutions to understand the reasons behind models' behaviours and the information they extract, to be able to evaluate their potential issues (accuracy, fairness, biases).


  • This thesis will investigate methods design to explain machine learning systems typically used in NLP while integrating an interactive process with the system users.


  • **Thesis subject**: Modern machine-learning based AI systems, while achieving good results on a lot of tasks, still appear as "black-box" models, where it is difficult to trace the path from the input (a text, an image, a set of sensor measures) to the decision (classification of a document, an image, a situation).


  • The issue of explainability poses two different problems: (1) what is a good explanation, and specifically what is a good explanation in the context of textual models? and (2) how to scale existing explanation methods to the kind of models used in NLP tasks?


  • About (1), existing methods for image classification or tabular data tend to rely on the extraction of a set of pixels or features that are sufficient for generating predictions, or increase the probabilities of the prediction.


  • It is less straightforward for textual input, which consists of words, but whose meanings are inter-related in a given context (for instance "good" in a review could be an indication that the review is positive ... unless it is preceded by "not").


  • So the first problem of this thesis will be to provide humanly acceptable explanations of simple text classifiers such as those foreseen for the detection tasks in the dedicated sub-project of Descartes.


  • About (2), modern NLP models are based on very large and complex architectures, such as the transformer family. Logically sufficient or causally satisfying explanations are difficult to get for such cases, as both such methods suffer from scalability problems.


  • So we will explore heuristics based on our solution to the first problem guiding an interactive procedure between explainee (the person requesting the explanation) and the ML system whose predictions should be explained. We will evaluate the procedure on those users targeted for the use cases of the project. Brian Lim from NUS Singapore will help design the validating experiments.


  • **Competences for the student**: A background in Computer Science and/or Machine learning. Familiarity or a willingness to acquire a familiarity with both model based and model agnostic explanation paradigms that use either logical or statistical methods.


  • A familiarity with NLP / dialogue would be a plus. Given the nature of the project, the student should be open to work in a cross-disciplinary environment, and have good English communication skills


  • **Supervision**: The thesis will happen within the France-Singapore collaboration, with advisors from both sides. The student will be registered at the University of Toulouse, and part of the IRIT lab, but is expected to spend a good part of the thesis in Singapore at the partner lab, with funding provided by the Descartes project.


  • The thesis will be supervised on the French side by Nicholas Asher and Philippe Muller, both NLP experts on text and conversation analysis, and co-advised by Nancy Chen from the A* lab, expert in NLP and dialogue, and Brian Limat the National University of Singapore, an expert on Human-Computer interaction. The French advisors will also spend time at NUS during the thesis.


  • Contact: [email protected], [email protected], [email protected]


These offer: Diving of complex terms for the extraction of information and the classification of clinical texts, LISN, Universite Paris-Saclay



  • Pierre Zweigenbaum [email protected]


  • Thesis in automatic language processing: Diving complex terms for information extraction and classification of clinical texts, LISN, Université Paris-Saclay


  • The LISN ILES team offers thesis funding as part of the ANR PREDHIC project.


  • Start date: February 2022 Location of the thesis: LISN, CNRS, Université Paris-Saclay Co-supervision: Pierre Zweigenbaum (LISN, Orsay), Emmanuel Morin (LS2N, Nantes) The overall goal of the thesis is to design a neural architecture that optimizes a text classification task based on the feature detection.


  • As complex terms play an important role in specialized fields, the thesis hypothesizes that better consideration of their representation will improve the detection of entities and the classification of texts. The field of application of the thesis is medicine.


  • Specifically, the classification task is the prediction of the rehospitalization or death of heart failure patients from the text and structured data of their electronic patient records.


Three Deep learning and Computer Vision Postdoc Positions Jointly Supervised in Aristotle University of Thessaloniki, Greece and Henan University, China



  • Three Deep learning and Computer Vision Postdoc Positions Jointly Supervised in Aristotle University of Thessaloniki, Greece and Henan University, China


  • Position Description Three postdoctoral positions are immediately available at the Data Science and Artificial Intelligence Team/Lab, Henan University; the postdoc fellows will be co-supervised by Prof. Ioannis Pitas from Aristotle University of Thessaloniki, Greece.


  • The postdoc fellows can stay 4-5 months with Prof. Ioannis Pitas at Aristotle University of Thessaloniki, Greece; but they should stay in China for at least 7-8 months per year.


  • Position Location The postdocs will be paid by Henan University, China. They will stay in Aristotle University of Thessaloniki (4-5 months), Thessaloniki, Greece, and Henan University (7-8 months), Kaifeng, China


  • Subject Areas Deep learning and Computer Vision, including but not limited to Object Detection/Recognition, Region Segmentation and related topics.


  • Salary 1800~2200 EUR net per month.


  • Application Deadline December 31st, 2022


  • Duration of the contract The duration of the contract is 24~36 months, and can be extended up to 48 months.


  • Salary a) The monthly net income of the Postdoctoral researcher is around 1800~2200 EUR (after taxes being deducted), paid by Henan University.


  • b) Additionally, the University will provide a total amount of 100,000 RMB as the candidate's research funding, which can be used for research purposes such as conferences, traveling, meetings, etc. Note: This research funding is once for all, not per year.


  • Note: The Living expenses in Kaifeng is relatively low. For instance,


  • Accommodation cost (two bedrooms, one living room) in Kaifeng is about 150~200 EUR per month. Food cost in KaiFeng is about 100-150 EUR per month.


  • Requirements/Qualifications 1. The successful candidates should hold a PhD degree in machine learning, or computer vision/pattern recognition, or other related fields.


  • 2. At the time of appointment, the applicants must be no more than 35 years old in age.


  • 3. The applicants should have good publication records in the above-mentioned fields.


  • About Aristotle University of Thessaloniki, Greece


  • Artificial Intelligence and Information Analysis Lab (https://aiia.csd.auth.gr/) Aristotle at University of Thessaloniki (AUTH) has one the best R&D records in Europe (72+R&D projects, mostly EC funded).


  • The Department of Informatics at AUTH is ranked 1st among the Greek Universities and 106th internationally in the field of Mathematics & Computer Science for 2019 in the Leiden Ranking list, which mainly depends on the scientific impact and staff publications.


  • It is also ranked 1st among the Greek Universities in the international ranking list Guide2Research, for 2021, having 5 faculty members in the list of top scientists in the scientific area of Computer Science and Electronics.


  • Aristotle University of Thessaloniki (AUTH), established in 1925, is among the most prestigious universities in Greece. Its computer science discipline is ranked #335 in the world in the US News University Ranking 2021.


  • Prof. Ioannis Pitas (34100+ citations, h-index 87) is an IEEE Fellow, EURASIP fellow and a world-renowned professor in multimedia systems and computer vision.


  • Homepage: https://aiia.csd.auth.gr/computer-vision-machine-learning/#people


  • https://scholar.google.gr/citations?user=lWmGADwAAAAJ&hl=en


  • About Henan University and KaiFeng, China


  • Henan University has a history of 109 years, and has been selected as the ''Double First Class'' University by the Ministry of Education of China, see http://en.henu.edu.cn/.


  • KaiFeng is a historical city (capital of Song Dynasty) and a famous tourist destination, see


  • https://en.wikipedia.org/wiki/Kaifeng.


  • How to apply


  • Applications are only accepted via Email.


  • All the following documents should be sent to Prof. Ioannis Pitas and Prof. Chongsheng Zhang simultaneously, at [email protected] and [email protected], with the email subject ''Postdoc application'':


  • - 1. A cover letter


  • - 2. Full CV and list of publications


  • - 3. Two reference letters.


Post-doctoral position at CIAD Laboratory: Hybrid Artificial Intelligence in Plasmonic



  • This project aims at developing compact all-optical Arithmetic and Logic Units (ALU) exploiting the spatial and spectral distributions of 2D confined plasmons modes in planar cavities tailored in ultrathin Au or Ag crystals. Yet, the optimization of the logic gate output contrast, the definition of the logic function reconfiguration schemes and the generalization of this concept towards complex ALU is a non-intuitive challenge. DALHAI addresses the ALU design challenge with a four-stage strategy that relies, first, on the mode symmetry considerations that led to the successful numerical and experimental results obtained on the DH devices. Second, evolutionary optimization will be implemented to efficiently survey the parameter space. Yet the discovery of complex ALU configurations will be limited by the intuitive starting points. Third, to overcome this limitation, DALHAI will develop powerful Hybrid Artificial Intelligence (HAI) tools and interface them with optical simulations and experimental data. Fourth, once trained, the HAI will propose device geometries and excitation protocols to solve the inverse design of complex reconfigurable ALUs. Nanofabrication, simulations, optical benchmarking, operation and reconfiguration of HAI-proposed ALUs will be performed. The experimental fabrication and optical testing and the numerical simulations of plasmonic ALUs will be performed by CEMES (CNRS, Toulouse) and ICB (CNRS, Dijon). CIAD (Univ. Bourgogne, Dijon) develops the HAI in strong interaction with all partners.


  • Topic: Generative Antagonist Network (GAN – Deep Learning), Knowledge and reasoning Engineering (OWL-DL), Explainable Artificial Intelligence *


  • DALHAI will adapt Hybrid Artificial Intelligence (combination of induction and deduction reasoning processes) to assist the design of the complex ALU to step up in complexity, numbers of input/output, and reconfigurability beyond intuitive design. DALHAI ambitions to enhance the innovation capacity by merging interdisciplinary fields and to establish national and European leadership in HAI-reinforced nano-photonics. In this regard, DALHAI aims at software maturity at TRL7. In this project, we have already built the first proposition of antagonistic reasoning allowing the prediction of the photon. These results are very encouraging and they were obtained quickly enough to validate the approach described in the initial file. These results open a new and very innovative artificial reasoning path that we wish to explore with this postdoc position. The mission is spread over several phases. 1/ The first phase revolves around the AI ​​approach put in place. This phase consists of analyzing the limits of simulation tools in the field of plasmonics and nanophotonics and familiarizing oneself with the IA approach implemented. 2/ The second phase aims to automate the generation of a synthetic data corpus by implementing the hybridization of artificial intelligence algorithms. 3/ The third phase of this post-doctoral position consists in setting up symbolic and connectionist artificial intelligence approaches to solve problems applied to the field of nano-plasmonics such as the prediction of optical signals, the optimization of the efficiency of logic gates


  • Starting period: February 2022 – duration of 12 months (can be extended to 18 months according to applicant profile).


  • Localization : Laboratoire CIAD – Université de Bourgogne, 64 rue de Sully, 21000 Dijon


  • Salary average 2500€ a gross monthly depending on qualifications and situation.


  • Applicants are required to have a PhD in Computer Science, a strong background in Machine Learning/Deep Learning (GAN), and eventually semantic web technologies. Fluency in written/spoken English is required too. A good publication record and strong programming skills will be a plus. Applications will be accepted until the position is closed. Applicants should send a full CV including a complete list of publications, a cover letter indicating their research interests, achievements to date and vision for the future, as well as either support letters or the name of 2 persons that have worked with them.


  • Contact: Christophe Nicolle ([email protected])


Postdoc position in Strasbourg: DL, Domain Adaptation, Multi-Modal Representations



  • A Postdoc position is open at University of Strasbourg (ICube lab) - France


  • Deep Learning, Domain Adaptation, Multi-Modal Representations


  • The position will be funded for two years (initially for one year, renewable for an additional year). The candidate will join the SDC research team under the supervision of Dr Thomas Lampert, the Chair of Data Science and Artificial Intelligence, and join his international team of PhD students and engineer to develop novel deep learning approaches to domain invariant representation learning (particularly in multi-modal data), with application (but not restricted) to Medical Imaging and Remote Sensing. The funding is not connected to a particular project, so it is the perfect opportunity for a strong candidate to explore new directions under the supervision of the Chair.


  • The successful candidate will have (or will soon obtain) a PhD in computer science or related domain and have experience in deep learning and applied machine learning and a strong level of written and spoken English. Experience with transformers, GANs, autoencoders, and/or unsupervised/self-supervised DL (autoencoders, etc) would be a plus. You will join a growing team and will have the freedom to follow your interests in a direction complementary to the abovementioned research focusses. You will be expected to target leading outlets in the field of machine learning and a strong track record in CVPR/ICCV/ECCV, NIPS/ICML/ICLR, or PAMI/IJCV/TIP. Candidates who are able to carry out the highest quality research independently, to co-supervise PhD students, and to give their input on a number of projects being carried out in the team are pursued. You will have access to state-of-the-art hardware for deep learning.


  • Send a letter of motivation, your CV, and an example publication to Thomas Lamper and Gisèle Burgart ([email protected] and [email protected] - !remove the numbers!) with the subject beginning with [Chaire Postdoc].


  • The position will remain open until a suitable candidate is found and the starting date will be agreed upon with the successful candidate (but can start ASAP).


  • Detailed Description: https://seafile.unistra.fr/f/5931f91dcffb401db566/?dl=1


Internship offer : Structural attention mechanism for Graph Transformers



  • Interested in Deep Learning and Graph ? Apply to our Master internship position in Structural attention mechanism for Graph Transformers.


  • As part of the French National Project called ANR CodeGNN, you will have the opportunity to continue with a funded PhD in our team !


  • More information below or at : https://lifat.univ-tours.fr/medias/fichier/lifat-internship-anr-codegnnen_1637256648035-pdf


  • Keywords Transformers; Graphs, Machine learning


  • Most of the objects of interest of our today's life are based on discrete objects with sequential (strings) or more complex (graph) relationships. We can evoke the relationships between people in social graphs, the bounds between atoms in a molecule or the topographic distance between speed sensors in traffic analysis, to name a few.


  • The prediction of the properties of such objects falls in the scope of structural pattern recognition.


  • A first breakthrough in this field has been provided by the introduction of Graph Neural Networks (GNNs). As graph kernels, these networks provide a strong connection between graphs and machine leaning techniques. Moreover, as other deep learning techniques, GNNs avoid handcrafting the design of a similarity measure between graphs. GNNs are based on two operations, namely, Graph convolution and Graph decimation/pooling.


  • A second breakthrough in this field has been provided by Transformer that were adapted to graph. The main adaptations of the transformers to the graph domains are : 1°) The attention mechanism is a function of the neighborhood connectivity for each node in the graph. 2°) Second, the positional encoding that provides a structural context for each node in the graph. 3°) The architecture is extended to edge feature representation.


  • Internship objectives and tasks : Including structural information in the attention mechanism


  • In previous works, a first GNN architecture working directly into graph space was proposed. Convolution and pooling operators are defined in graph domain while allowing the use a back-propagation algorithm during the learning step. Especially, the convolution is replaced by a graph matching solver applied on a subgraph rooted around each node of the graph.


  • The idea to be developed in the master thesis is to investigate the use of a graph matching solver in the objective to a structural attention mechanism.


  • The aim of this master thesis is to :


  • 1. Study alternative computation of attention mechanism to take into structural information.


  • 2. Propose a graph transformer model based on structural attention mechanism.


  • 3. Program these models (in Python), and compare them to the state-of-the-art on standard datasets for different applications.


  • This internship is proposed in the context of a French National Project (ANR) called CodeGNN.


  • Supervisors


  • Jean-Yves Ramel (University of Tours, France) [email protected]


  • Romain Raveaux (University of Tours, France) [email protected]


  • Prerequisites:


  • Bachelor degree in Computer Science, Applied Mathematics, Data Science, or similar.


  • Skills (with experiences): deep learning, Python programming, numerical analysis.


  • When, where and how much


  • When : The internship will take place between March until September 2022 for 5 months. This position is limited to 5 months but it could be continued by a PhD position. Funding for a PhD have been already obtained.


  • How much : The internship will be granted with around 500 euros/month (legal allowances).


  • The internship will take place at the Computer science laboratory of Tours / Laboratoire d'Informatique Fondamentale et Appliquées de Tours (LIFAT, http://lifat.univ-tours.fr), Tours, France


  • How to apply


  • Please submit pdf files of your CV to: [email protected] and [email protected]. If your application is selected, you will then be contacted for further information and interview details.


Openings in Université Côte d'Azur (Inria/CNRS I3S) on multimedia processing, immersive systems, and ML



  • Dear all, We are currently seeking motivated candidates for multiple job openings in the Inria and CNRS I3S laboratories of Université Côte d'Azur on various topics in multimedia analysis and processing, immersive systems, and user motion modeling. Please consult the following links for detailed post descriptions and application information:


  • - Engineer in computer vision and data processing (https://recrutement.inria.fr/public/classic/en/offres/2021-04120)


  • - Engineer in immersive multimedia systems (https://www.i3s.univ-cotedazur.fr/~sassatelli/files/Engineer_position_UCA_AI4Media.pdf)


  • PhD thesis in multimedia analysis (https://www.i3s.univ-cotedazur.fr/~sassatelli/files/PhDposition_IRIT-I3S_ANR_TRACTIVE.pdf)


  • Modeling 6 DoF human motion data in extended reality (https://www-sop.inria.fr/members/Hui-Yin.Wu/offers/stage_creattive3d_6dof_modeling.pdf)


  • Image and text processing for multimodal film analysis (https://www-sop.inria.fr/members/Hui-Yin.Wu/offers/stage_tractive_data.pdf)


  • Hui-Yin Wu (Helen) Research scientist - Inria Starting Faculty Position


  • Biovision Team @ Centre Inria d'Université Côte d'Azur + 33 4 92 38 79 28


Job: Postdoc (18 months), Computational Diachronic Semantics (LIPN Univ. Sorbonne Paris Nord, Labex EFL)



  • Postdoctoral researcher in Computational Diachronic Semantics


  • Labex EFL (Empirical Foundations of Linguistics, Paris, https://www.labex-efl.fr/)


  • Strand 5: Computational Semantic Analysis


  • Research area: interpretable computational models for automatic detection and monitoring of semantic evolutions: combination of Contextual Embeddings and Pattern Mining approaches


  • Contract duration: 18 months


  • Location: Paris


  • Research Laboratory: Sorbonne Paris Nord University, LIPN UMR7030 CNRS


  • Application deadline: January 15, 2022


  • Audition period: January 15-30, 2022


  • Job Starting date: from February 1, 2022


  • Context, Issues and research axes


  • Languages ​​are constantly evolving, driven by the need to adapt to socio-cultural and technological developments and to make communication more efficient and expressive. In particular, new words are forged or borrowed from other languages, some words become obsolete, others acquire new meanings or lose existing meanings.


  • In NLP, the study of language dynamics, especially from the lexical point of view, has gained audience in recent years, complementing synchronic approaches. The field of research is structuring itself, with recent state of the art (Monteirol et al., 2021; Tahmasebi et al., 2021) and several scientific events (International Workshop on Computational Approaches to Historical Language Change 2019 and 2021, ACL 2019 and 2020). Two initial evaluation tasks have been proposed (Unsupervised Lexical Semantic Change Detection Task, SemEval2020) and reference sets have been set up for four languages ​​(English, Latin, Swedish and German).


  • Lexical change detection systems have followed advances in NLP methods: after the first systems essentially based on frequency changes (for example Gulordova & Baroni, 2011), systems used word embeddings (Kim et al., 2014, Schletchweg et al., 2019) and more recently contextual embeddings (Hu et al., 2019; Martinc et al., 2019; Giulianelli et al., 2020). These latter systems generally proceed by grouping the contextual vector representations of the different uses into clusters of meaning, then detect changes according to different metrics (Monteirol et al. 2021). Current systems still face many limitations. Mainly, the opacity of neural models does not make it possible to characterize these evolutions, in particular it is difficult, if not impossible, to link the semantic changes to linguistic morphological, syntactic or lexico-syntactic features, or to categorize the types of changes (extension, restriction, metaphor, metonymy, etc.). To this end, one avenue would be to combine neural approaches with Pattern Mining (Béchet et al. 2015) or collocation extraction approaches from corpus linguistics (for example Gries, 2012) which make it possible to extract the most salient lexico-syntactic patterns of a given meaning from a corpus of occurrences and thus identify the evolution. It would also be interesting to use the contextual information of the occurrences (date, type of source, domain, diatopic and diastratic features, etc.) to characterize and follow the evolution of usages.


  • The job main objective is therefore to set up a system combining these approaches to allow an automatic characterization of semantic evolutions. The first step will consist in experimenting with state-of-the-art models for detecting changes. The second step will then try to combine contextual embeddings and pattern mining approaches / collocation extraction to highlight the linguistic characteristics of each of the meaning clusters and their evolution. The studied corpora will be mainly in English and French. The postdoctoral fellow will work in collaboration with computer scientists and linguists from the Labex who are currently building a reference corpus of semantic evolutions for French (following the Durel methodology: Schlechtweg et al., 2018).


  • Other issues may also be addressed by the recruited person, and in particular: current systems do not take into account the graduality of evolutions, generally being limited to comparing two synchronic language states; to get the vector representation of a lexis in a context, it is possible to use one of the hidden layers or a combination of them.


  • There is currently no consensus on the most adequate layer to take into account to obtain the most adequate semantic representation.


  • The recruited person will join the strand 5 (“Computational Semantics”) of the Labex, specifically the research team working on the “Semantic Variation and Change” operation which aims to:


  • - develop new models and methods for the automatic detection of lexical semantic changes, the typology of changes from intra- and extra-linguistic points of view;


  • - develop a reference dataset of semantic evolutions in contemporary French, based on available diachronic corpora.


  • Candidate profile


  • - PhD in computer science specialised in Computational Linguistics and Machine Learning


  • - deep learning methods and language models attested training and experience


  • - working language: French and / or English


  • Application


  • Please send :


  • - a cover letter


  • - a description of the research project related to the research questions


  • - a CV with a list of publications and 3 representative publications (pdf or link),


  • - letters of recommendation or names of two referees.


  • to [email protected] and [email protected] before January 15, 2022.


  • The auditions of the pre-selected candidates will take place at the end of January 2022.


Job : Postdoc (18 months), Deep learning for opinion mining in human testimonials related to industrial accidents, LITIS University of Rouen



  • The Machine Learning team of the LITIS laboratory is still looking for a post-doctoral researcher or a research engineer, for a position in deep learning for sentiment analysis, funded by by the ANR CATCH project.


  • Title: *Deep Learning for opinion mining in human testimonials related to industrial accident*


  • Keywords: deep learning, NLP, sentiment analysis, twitter


  • Location: LITIS lab., University or Rouen Normandy, Rouen, France


  • Duration: 18-months, starting as soon as possible


  • Salary: ~2300€/month (before income tax), including social security coverage in line with French regulations


  • Please find the detailed description of the position below. This position is still unfilled, so please feel free to forward to anyone who might be interested.


  • Simon Bernard LITIS / NormaSTIC FR CNRS 3638


  • Université de Rouen Normandie, France http://pagesperso.litislab.fr/sbernard/


  • Post-doctoral or Research Engineer position: Deep Learning for opinion mining in human testimonials related to industrial accident*


  • The Machine Learning team at the LITIS laboratory, the computer science laboratory of the University of Rouen Normandy, is looking for a post-doctoral researcher or a research engineer on a 18-months contract, starting as soon as possible. The position will be financed by the ANR research project CATCH (french acronym for "Automatic Understanding of Human Sensors Testimonials"), which involves the R&D center of the company Saagie, specialized in B2B DataOps solutions, Atmo Normandie, one of the approved French air quality monitoring associations and LITIS.


  • Scientific context:*


  • The ambition of the CATCH project is to propose artificial intelligence and deep learning tools to take into account and automatically exploit the multitude of human testimonies related to an industrial accident and its consequences on the environment and health. By involving the population in the collection and analysis of data, particularly through social networks, and by providing effective means for interpreting this data, the proposed solution should contribute to providing answers to the worrying problem of industrial accidents and their consequences.


  • The overall objective is first to draw up a precise cartography of the nuisances in order to follow the propagation and the evolution of the phenomena in time, and then to analyze and characterize the sentiment of the population and its evolution throughout the crisis. To do so, we intend to exploit testimonials collected on the ODO platform of Atmo Normandie, which combines these testimonies with geographical information, in conjonction with data extracted from the micro-blogging platform Twitter. Since these data are primarily texts, state-of-the-art approaches from the Natural Language Processing (NLP) field are favored, in particular, self-supervised deep learning methods such as Transformers that are known to be the most performant today for a wide range of NLP tasks.


  • *Research goals:*


  • The objective of this research work is twofold:


  • 1. The automatic generation of a map of nuisances around the site of an industrial incident to monitor the propagation and the evolution of the phenomena over time.


  • 2. The automatic recognition of the population's perception and its evolution throughout the crisis.


  • Related to these tasks, the post-doctoral researcher or the research engineer will be in charge of proposing and implementing solution for:


  • - extracting and linking twitter data with testimonials from the ODO dataset, which is fully labelled and associates textual testimonies with geographical data. The interest in establishing this link is to be able to enrich the data from the ODO platform to refine the mapping of nuisances in real time. This could be achieved for example, by using pseudo-labelling techniques or Constrative Representation Learning methods which have recently been applied to text data.


  • - detecting in all the testimonials collected from Twitter or from the ODO platform, the presence (or absence) of several pre-identified emotions (e.g. surprise, fear, anger, sadness, disgust, etc.), several of which can be expressed at the same time.


  • This work will therefore involve being familiar with the state-of-the-art NLP deep learning methods and in particular with their applications to sentiment analysis and opinion mining tasks. It will also require experience with the use and exploitation of data from Twitter in a data science context.


  • *Application:*


  • The successful applicant will:


  • - have a PhD or a Master's degree or an engineering degree in computer science or applied


  • - mathematics, with a specialization in machine learning or data mining.


  • - have strong programming skills and in-depth understanding of statistics and machine learning.


  • - have already worked with deep learning architecture dedicated to texts (RNNs, Transformers, etc.) and/or images (CNNs, FCNs, GANs).


  • Your application should include:


  • - curriculum vitae


  • - statement of past experiences with deep learning and/or NLP techniques


  • - representative published articles where applicable


  • - contact information for the referent PhD supervisor or last diploma's referent teachers


  • Application must be sent to :


  • - Simon BERNARD, University of Rouen Normandy, [email protected]


  • - Clément CHATELAIN, INSA Rouen Normandy, clement.chatelain@insa-rouen. fr


  • - Alexandre PAUCHET, INSA Rouen Normandy, alexandre.pauchet@insa-rouen. fr


Job : CDI, TAL expert, DGA MI, Bruz (Rennes)



  • Hello everyone, The Directorate General of Armaments (DGA, Directorate of the Ministry of the Armed Forces) is recruiting on a permanent contract an expert in Automatic Language Processing, description of the offer below.


  • Do not hesitate to contact us for more information.


  • Place of Work: Bruz (35) Type of employment: Full-time permanent contract Job level: Engineer Profession: Artificial Intelligence / Natural Language Processing Salary: The salary will be determined according to professional experience Desired start date: March 2022


  • Description of the entity:


  • DGA-MI The primary mission of the Directorate General of Armaments (DGA) is to provide the armed forces, the services of the Ministry of the Armed Forces and certain interministerial services with the systems and tools necessary for their missions.


  • Within the DGA, the Information Control Center – DGA MI provides technical expertise for all weapons programs.


  • The DGA MI is located in Bruz, on a pleasant site and close to Rennes (arrangements to come by car / public transport / bike), and offers a privileged environment allowing a very good personal and professional balance.


  • In this context, we are recruiting a full-time permanent engineer specialized in automatic language processing capable of providing and promoting technical expertise in this field in the context of projects that meet major scientific and technical barriers: specialty data, rare and poorly endowed languages, etc.


  • You will join a team composed of experts working from transversal way and in close collaboration with other departments.


  • Missions: You work as a technical expert in artificial intelligence applied to NLP for:


  • - The collection of operational needs and the upstream specification for the development of new systems;


  • - The follow-up of the realization by the industrialists and the qualification of systems for the military operational needs;


  • - The establishment and maintenance of the means of evaluation of TAL systems, including the design and implementation of algorithms for language data processing and evaluation of NLP technologies;


  • - The realization of proofs of concept; - Missions related to open innovation, as well as technology watch and ecosystem monitoring activities (academics, start-ups, SMEs, large groups).


  • Your expertise will focus on the following areas: - Tools and technologies related to the exploitation of structured and/or unstructured multilingual language data;


  • - Valorization of complex data and visualization of information;


  • - Analysis, training and evaluation of learning models.


  • As part of this expertise, you will be required to cover the following applications: - Classification, sorting and filtering of documents - Comprehension aid - Decision-making assistance - Machine translation - Text mining and information extraction


  • - Analysis of social media - Analysis of emotions, opinions, feelings, intentions


  • - Advanced search engines


  • - Human-Machine dialogue in natural language - Help with modeling and representation of knowledge Desired profile:


  • You hold at least a diploma = F4me Bac + 5 (engineer, master, doctorate) in computer science, artificial intelligence or data science, with a specialty in natural language processing. A minimum of 3 years' professional experience is highly desirable.


  • You are a specialist in learning methods, including: - Machine, statistical and symbolic learning; - Deep neural networks, convolutional networks. Rigorous, you have a good spirit of synthesis.


  • You also demonstrate good interpersonal skills and enjoy teamwork.


  • Application: Application (CV in French) to be submitted to [email protected] ouv.fr indicating in the subject the reference 2022-ASC/EORD_TALIA.


  • Authorization: As the position requires access to information relating to the secrecy of national defense, the holder will be subject to an authorization procedure at the Top Secret level, in accordance with the provisions of articles R.2311-1 and following of the Defense Code and the IGI 1300 SGDSN/PSE of November 30, 2011.


  • Camille DUTREY Expert in Natural Language Processing DGA DT / MI / SDTI / ASC2 / EORD 136, La Roche Marguerite BP 7 – 35998 RENNES Cedex 9 Tel.: 02 99 42 91 47 – 862 357 9147


Job offer at CEA LIST



  • Hello I belong to the Vision and Learning Laboratory for Scene Analysis (LVA) of CEA LIST.


  • We are looking for a researcher or engineer in vision and deep learning for the development of AI and vision technologies for very high performance sports. The contract is a fixed-term contract of 18 to 36 months.


  • If you are interested, you will find attached a detailed description of the position. Do not hesitate to contact us by email with your CV and a cover letter.


  • Sincerely, Bertrand LUVISON Research Engineer


  • CEA LIST DIASI/SIALV/LION


  • CEA SACLAY – NANO INNOV – Bât. 861 – Mail point 184


  • 91191 Gif-sur-Yvette Cedex


  • – France Tel : 01 69 08 01 37 www-list.cea.fr


Post-doctoral or Research Engineer position



  • Deep Learning for opinion mining in human testimonials related to industrial accident


  • Location : LITIS lab., University of Rouen Normandy, Rouen, France


  • Duration : 18-months, starting as soon as possible


  • Salary : ~2300€ / month (before income tax), including social security coverage in line with French regulations


  • The Machine Learning team at the LITIS laboratory, the computer science laboratory of the University of Rouen Normandy, is looking for a post-doctoral researcher or a research engineer on a 18-months contract, starting as soon as possible. The position will be financed by the ANR research project CATCH (french acronym for "Automatic Understanding of Human Sensors Testimonials"), which involves the R&D center of the company Saagie1 , specialized in B2B DataOps solutions, Atmo Normandie2 , one of the approved French air quality monitoring associations and LITIS.


  • Keywords


  • Deep learning – Natural Langage Processing – Sentiment Analysis / Opinion Mining


  • Scientific context


  • The ambition of the CATCH project is to propose artificial intelligence and deep learning tools to take into account and automatically exploit the multitude of human testimonies related to an industrial accident and its consequences on the environment and health.


  • By involving the population in the collection and analysis of data, particularly through social networks, and by providing effective means for interpreting this data, the proposed solution should contribute to providing answers to the worrying problem of industrial accidents and their consequences.


  • The overall objective is first to draw up a precise cartography of the nuisances in order to follow the propagation and the evolution of the phenomena in time, and then to analyze and characterize the sentiment of the population and its evolution throughout the crisis.


  • To do so, we intend to exploit testimonials collected on the ODO platform3 of Atmo Normandie, which combines these testimonies with geographical information, in conjonction with data extracted from the micro-blogging platform Twitter.


  • Since these data are primarily texts, state-of-the-art approaches from the Natural Language Processing (NLP) field are favored, in particular, self-supervised deep learning methods such as Transformers4 that are known to be the most performant today for a wide range of NLP tasks5 .


  • Research goals


  • The objective of this research work is twofold:


  • The automatic generation of a map of nuisances around the site of an industrial incident to monitor the propagation and the evolution of the phenomena over time.


  • The automatic recognition of the population's perception and its evolution throughout the crisis.


  • Related to these tasks, the post-doctoral researcher or the research engineer will be in charge of proposing and implementing solution for:


  • extracting and linking twitter data with testimonials from the ODO dataset, which is fully labelled and associates textual testimonies with geographical data. The interest in establishing this link is to be able to enrich the data from the ODO platform to refine the mapping of nuisances in real time.


  • This could be achieved for example, by using pseudo-labelling techniques or Constrative Representation Learning methods which have recently been applied to text data


  • detecting in all the testimonials collected from Twitter or from the ODO platform, the presence (or absence) of several pre-identified emotions (e.g. surprise, fear, anger, sadness, disgust, etc.), several of which can be expressed at the same time.


  • This work will therefore involve being familiar with the state-of-the-art NLP deep learning methods and in particular with their applications to sentiment analysis and opinion mining tasks. It will also require experience with the use and exploitation of data from Twitter in a data science context.


  • Application


  • The successful applicant will:


  • • have a PhD or a Master's degree or an engineering degree in computer science or applied mathematics, with a specialization in machine learning or data mining.


  • • have strong programming skills and in-depth understanding of statistics and machine learning.


  • • have already worked with deep learning architecture dedicated to texts (RNNs, Transformers, etc.) and/or images (CNNs, FCNs, GANs).


  • Your application should include:


  • • curriculum vitae


  • • statement of past experiences with deep learning and/or NLP techniques


  • • representative published articles where applicable


  • • contact information for the referent PhD supervisor or last diploma's referent teachers


  • Contact


  • Application must be sent to :


  • • Simon BERNARD, University of Rouen Normandy, simon.bernar d @univ-rouen.fr


  • • Clément CHATELAIN, INSA Rouen Normandy, [email protected]


  • • Alexandre PAUCHET, INSA Rouen Normandy, [email protected]


Jobs - Academic positions at the University of Sheffield



  • Dear All We are currently recruiting for four exciting positions for Lecturer/ Senior Lecturer in the Department of Automatic Control and Systems Engineering at the University of Sheffield, United Kingdom.


  • We are looking for innovative researchers in the broad areas of Systems and Control, Artificial Intelligence and Autonomy and Complex Systems who can lead research of an international standard, contribute to the department’s undergraduate and postgraduate taught programmes and supervise PhD students.


  • If this is you, then please find out more and apply online jobs.ac.uk


  • Job Reference Number: UOS030936 Job Title: Lecturer/Senior Lecturer x 4 posts Faculty: Faculty of Engineering Department: Automatic Control and Systems Engineering Salary: Grade 8: £41,526 - £49,553 per annum. Potential to progress to £55,750 per annum through sustained exceptional contribution. Grade 9: £53,348 - £60,022 per annum. Potential to progress to £69,557 per annum through sustained exceptional contribution. Closing Date: 4th January 2022


  • Dawn Aykroyd


  • Department of Automatic Control & Systems Engineering


  • The University of Sheffield Amy Johnson Building, Mappin Street, Sheffield, S1 3JD


  • Telephone: +44 (0)114 222 5134


Postdoctoral offer in HMI and adaptive interfaces - NavalGroup / LabSticc



  • Dear colleagues, We are still actively looking for a candidate for the postdoctoral offer below, for a one-year project with Naval Group on the acceptability of adaptive human-machine systems.


  • Please disseminate this postdoctoral offer to your future ex-doctoral students or young doctors, and more generally in your networks.


  • Project Description


  • The progress made over the last ten years in terms of learning algorithms, both in terms of intrinsic performance and deployment, now makes it possible to envisage the implementation of systems with mechanisms toadapt to their environment and their users. This evolution particularly concerns sociotechnical systems in which users can now be observed and "understood" by the machine so that the latter adapts its own operation to the states (fatigue, stress, ...) or behaviors (recurrent errors, ...) detected.


  • While this system paradigm shift is now technically accessible, there is little work on how to make such a system, by flexible or versatile construction, acceptable to operators. All traditional ergonomics has indeed been based on the hypothesis of a system with stabilized operation and has been rather interested in the evolution or change of behavior of human operators, and few works have dealt with reactions of human operators to the flexible or dynamic behaviors of a learning system (both vis-à-vis its user but also its environment and the tasks to be accomplished).


  • This project aims to fill this gap by focusing on how self-adaptive interaction is understood, accepted and beneficial by the user (including on H/S performance) according to his "cognitive" state (stress, ...).


  • The key points of the study may include:


  • the impact of adaptive interaction on operator trust in the system


  • the impact of the levels of information transparency conveyed and the modalities of interaction on the acceptability of the system


  • and more generally the improvement in performance obtained with a system "increased in adaptability" compared to a system with fixed behavior.


  • The full description and other contact


  • information can be found here: https://institutminestelecom. recruitee.com/o/postdoctorant-acceptability-of-interactions-humans-systems-self-adaptives-modalities-of-implementation-to-optimize-interaction-cdd-24-months


  • The offer generally mentions all profiles covering human-machine interactions / ergonomics / cognitive psychology and profiles including AR / VR will be particularly considered.


  • The position is to be filled quickly, do not hesitate to contact me for more information.


Post-doctoral research position - L3i - La Rochelle France



  • Title : Extraction of graphic elements in comics books for emotion recognition


  • The L3i laboratory has one open post-doc position in computer science, in the specific field of document image analysis, pattern recognition, machine learning


  • Duration: 12 months (an extension of 12 months will be possible)


  • Position available from: As soon as possible, 2021


  • Salary: approximately 2100 € / month


  • Place: L3i lab, La RochelleUniversity , France


  • Specialty: Computer Science/ Image Processing/ Machine Learning / Document Analysis/ Pattern Recognition


  • Contact:Jean-Christophe BURIE (jcburie [at] univ-lr.fr)


  • Position Description


  • The L3i is a research lab of the University of La Rochelle. La Rochelle is a city in the south west of France on the Atlantic coast and is one of the most attractive and dynamic cities in France. The L3i works since several years on document analysis and has developed a well-known expertise in ‘Bande dessinée”, manga and comics analysis, indexing and understanding.


  • The work done by the post-doc will take part in the context of SAiL (Sequential Art Image Laboratory) a joint laboratory involving L3i and a private company. The objective is to create innovative tools to index and interact with digitised comics. The work will be done in a team of 10 researchers and engineers.


  • The work will consist in developing original approaches for extracting and recognizing graphics elements in comic panels in order to recognize emotions. Authors usually used different strategies for representing emotions such as shape of speech balloon, specific symbols, colour of the faces, etc. These elements are drawn among the other graphic elements (main characters, scenery, …) making the localisation and the extraction challenging. In order to extract these specific elements, the development of original approaches will be necessary. Deep learning-based strategies can be explored to reach this goal. This work will be done in collaboration with other researchers working on text understanding.


  • Qualifications


  • Candidates must have a completed PhD and a research experience in image processing and analysis, pattern recognition, machine learning/deep learning.


  • General Qualifications


  • • Good programming skills mastering at least one programming language like Python, Java, C/C++, deep learning frameworks


  • • Good teamwork skills


  • • Good writing skills and proficiency in written and spoken English or French Applications


  • Candidates should send a CV and a motivation letter to jcburie[at]univ-lr.fr.