PhD position - Dynamic community detection in node-attributed networks



  • PhD. position in Computer Science at Université Côte d'Azur I3S laboratory , https://www.i3s.unice.fr ===========================================================


  • Scientific description ----------------------


  • Many complex systems in nature and society can be described in terms of networks where the entities under study are represented by nodes, the links between these entities by edges, and the characteristics of the entities by attributes on each node, so called node-attributed graphs. The existence of a community structure is an important property of real-world networks because it often helps to explain the functionality of the system. Despite the ambiguity in the definition of community, many techniques have been developed for effective and efficient community detection, often considering the problem as clustering nodes in the graph. However, the algorithms generally rely on the use of only one source of information: the network structure. In recent years, community detection algorithms that also consider the attributes of the nodes have been proposed, with varying degrees of success (Chunaev, 2020). As community detection is NP-hard and an exhaustive search over all possible solutions is usually intractable, nature-inspired techniques have been employed and compared (Osaba et al., 2020) but on small or medium datasets (around 1000 nodes). Moreover, works mainly focused on the different criteria to combine in a multi-objective context and less on the representation of solution themselves. However, some progress was made in this area (Garza-Fabre et al., 2018; Luo et al., 2021).


  • On the opposite, some advances have been made in finding communities where structure of the graph is changing over time (Yin et al., 2021). It is often the case in social networks or protein-protein interactions in biology. However, techniques proposed in this context do not consider attributed graphs and often apply community detection on several static networks corresponding to temporal snapshots of the dynamic network. This approach generates communities that change too drastically over time, which does not reflect the reality.


  • The goal of this doctoral thesis is to tackle the problem of dynamic community detection in node-attributed networks by combining a bio-inspired algorithm with a local search. To do so, we envisage to use new insights (Tian et al., 2021) to scale up a first attempt we have made for active module detection in biology (Correa et al., 2019) and combine it with a greedy approach (Pasquier et al., 2021) in order to have better convergence capacity. This work will have concrete applications in the industrial project NewgenTOXiv that we have just started which aims to develop new in vitro toxicological tests.


  • Skills and profile


  • - Master degree in Computer Science, especially in Optimization and/or Data Mining - Very good programming skills are mandatory - Knowledge in bioinformatics could be a great help - Oral and written english is required. French is not mandatory


  • PhD supervision ---------------


  • - Claude Pasquier, CR CNRS HDR, http://claude.pasquier.net - Denis Pallez, Associate professor, http://denispallez.i3s.unice. fr


  • Applications should include the following documents in electronic format: - a cover letter, stating your motivation, scientific background, and research interests, - a detailed CV, - references of academics to be contacted (or recommendation letters).


  • Send all these documents by email to [email protected] and [email protected]


  • A pdf version of the position is available at https://claude.pasquier.net/data/thesis_2022.pdf


PhD position - Surgical Process Modelling with Graphical Event Models and Ontologies



  • PhD. position in Computer Science - Nantes Université


  • Surgical Process Modelling with Graphical Event Models and Ontologies


  • Supervisors : * Philippe Leray, LS2N, Nantes Université * Thomas Guyet, INRIA, Centre de Lyon * Pierre Jannin, LTSI, INSERM, Université Rennes 1


  • More details : https://uncloud.univ-nantes.fr/index.php/s/yffCR7p4G49T94s


  • Keywords : Artificial Intelligence, Probabilistic Graphical Event Model, Ontology, Machine Learning, Surgical Process Modelling.


  • Context DUKe (Data User Knowledge) research group at LS2N, UMR CNRS 6004, is one of the laboratory's main teams in "Data and Decision Science" field, with its skills in data manipulation, data mining and interaction. Within this framework, the research group has, among other things, developed numerous algorithms for learning and manipulating probabilistic graphical models (Bayesian networks, dynamic Bayesian networks, relational Bayesian networks, graphical event models) gathered within the PILGRIM C++ software library. This PhD thesis is part of the SPARS project (Sequential Pattern Analysis in Robotic Surgery: Understanding Surgery), funded by Labex CominLabs, in collaboration with LTSI/INSERM/Université Rennes 1 and INRIA.


  • The objective of this project is to propose data analysis methods to better understand complex technical human activities, such as surgery. Surgery is a complex activity, that depends on many factors, including the patient and surgeon characteristics. Such complexity and variability explain why there is almost no detailed study of the surgical practice yet. Until now, the surgical procedure performed in the operating room is considered as a whole, as a black-box and is technically described with few words. Analysis usually consisted in comparing impact of different surgical approaches or of different pre-operative clinical patient’s parameters on post-operative outcomes. In the SPARS project, we will rely on a combination of data and model-driven approaches to analyze and compare kinematics of whole surgical procedures acquired during robotic assisted hysterectomies.


  • Funding: The PhD fellowship is funded for 3 years from september-October 2022.


  • Profile of the candidate: The candidate should have a master's degree in computer science or equivalent, as well as knowledge of machine learning, probabilistic graphical models and knowledge representation. Good skills in machine learning is mandatory. Some knowledge in knowledge representation will be a plus.


  • The programming environment associated with this project also requires some knowledge of C++ programming language. The personal qualities expected are mainly autonomy and a taste for interdisciplinary work, rigour and abstraction, as well as writing skills (in French and English).


  • Application instructions: The application file should contain the following documents:


  • a curriculum vitæ (CV);


  • the official academic transcripts of all the candidate’s higher education degrees (BSc, License, MSc, Master’s degree, Engineer degree, etc.). If the candidate is currently finishing a Master’s degree, s/he must send the transcript of the grades obtained so far, with the rank among her/his peers, and the list of classes taken during the last year;


  • some recommendation letters (quality is more important than quantity, there);


  • * and a motivation letter written specifically for this position.


  • Send all of these documents by email to [email protected], [email protected] and [email protected]


18-month postdoc positions located in Nantes (France)



  • The RAMBO team (https://www.imt-atlantique.fr/en/research-innovation/teams/research-team-rambo) of IMT Atlantique, in collaboration with Smart Macadam ( https://www.smartmacadam.com/), is looking for candidates for two 18-month postdoc positions located in Nantes (France) on the following topics:


  • 1. Post-doctoral fellow speech recognition, artificial intelligence 18 month contract


  • French description: https://institutminestelecom.recruitee.com/l/fr/o/postdoctor ante-ou-postdoctorant-reconnaissance-automatique-de-la-parole-intelligence-artificielle-cdd-18-mois


  • English description: https://institutminestelecom.recruitee.com/l/en/o/postdoctorante-ou-postdoctorant-reconnaissance-automatique-de-la-parole-intelligence-artificielle-cdd-18-mois


  • 2. Post-doctoral fellow sound recognition, artificial intelligence 18 month contract


  • French description: https://institutminestelecom.r ecruitee.com/l/fr/o/postdoctor ante-ou-postdoctorant-reconnaissance-de-sons-intelligence-artificielle-cdd-18-mois


  • English description: https://institutminestelecom.r ecruitee.com/l/en/o/postdoctor ante-ou-postdoctorant-reconnaissance-de-sons-intelligence-artificielle-cdd-18-mois


  • For more details about the positions and to submit your application, please follow the links above.


  • Best regards, Mihai Andries Christophe Lohr


PhD position or postdoctoral topic in the framework of the CREAM project



  • Location: Orléans/Grenoble, France


  • Contacts: Emmanuel Schang (emmanuel.schang@univ-orleans. fr), Benjamin Lecouteux (benjamin.lecouteux@univ-greno ble-alpes.fr)


  • *Thesis or postdoctoral topic in the framework of the CREAM project :*


  • We are looking for a candidate for a thesis (or post-doctoral contract ) in Language Sciences / Computer Science on the theme of automatic speech processing.


  • The work will be carried out at the Laboratoire Ligérien de Linguistique (LLL, UMR 7270) and at the LIG-GETALP (Grenoble). Funding will be provided through the ANR CREAM (Machine-Assisted CREoles Language Documentation) project (https://sites.google.com/view/creamproject/home).


  • Creole languages, automatic speech processing, keyword detection, bilingual alignment, creole languages, speech processing, keyword spotting, bilingual alignment.


  • The CREAM project aims to provide linguists working on Creole languages with innovative tools in the collection and processing of oral data on languages with few resources. In the particular context of diglossia that often characterizes the Creole speaking space, the passage through the stage of corpus transcription is frequently perceived as a difficulty by field linguists . One consequence is the lack of available corpora.


  • The aim of this project is to pave the way for innovative methods of linguistic documentation and the creation of resources on Creole languages. Using state-of-the-art machine learning technologies, we seek to change the way language documentation is implemented in terms of building language resources and processing spoken corpora.


  • The focus will be on three main tasks:


  • automatic transcription in a language scenario with few resources,


  • Query-by-example: search for similar segments in Creole language corpora ,


  • Automated bilingual alignment between speech segments in a Creole language and a related language (French, English, Portugai s, following the Creoles).


  • Candidates will have a master's degree (or doctorate) in linguistics or computer science and will show a certain interest in automatic speech processing and so-called "rare" languages. Autonomy in python coding is essential, as well as basics in machine learning. M2 students who have supported before the end of September 2022 can apply for the doctoral contract. To apply for a post-doctoral contract, candidates will need to hold a doctoral thesis in computer science or linguistics (natural language processing ).


  • *Supervision* Emmanuel SCHANG (Doctor HDR in Language Sciences) Benjamin LECOUTEUX (Doctor HDR in Computer Science)


  • Applications must contain: CV + letter/motivation message + master's notes + letter(s) of recommendations; and be addressed to: Emmanuel Schang (emmanuel.schang@univ-orleans. fr), Benjamin Lecouteux (benjamin.lecouteux@univ-greno ble-alpes.fr).


PhD position (Paris, France) - SELEXINI project



  • https://selexini.lis-lab.fr/


  • Semi-supervised word sense and frame induction


  • - Contract duration: 36 months


  • - Starting date: October 2022 to December 2022


  • - Location: LLF laboratory, computational linguistics axis (http://www.llf.cnrs.fr/en/research-topics), Paris, France


  • - Advisors: Marie Candito (LLF laboratory, http://www.llf.cnrs.fr/en/research-topics) and Carlos Ramisch (LIS lab, TALEP team, https://talep.lis-lab.fr/)


  • - Net salary : 1750 € (including 64h teaching, optional)


  • - Application: The application file should be sent by May 9 to [email protected] and [email protected]. It should comprise:


  • - a CV (max 5 pages) with transcripts (Master), diplomas, internships


  • - a cover letter


  • - the names and contact of two referees


  • The candidates selected for interviews will send their Master thesis or other written work supporting their qualification for the project. They will be interviewed (remotely) between the end of May and mid-June 2022.


  • SELEXINI is a research project funded by the French National Research Agency (ANR) that focuses on semi-supervised word sense induction and semantic frame induction. The starting observation for this project is that identifying word meanings in context can lead to better performance and interpretability of NLP system predictions, but that the lack of large coverage sense-annotated data (coverage in terms of domains and of languages) hinders the use of lexicons in modern neural NLP.


  • The project aims at developing a word sense induction method by clustering occurrences, thus providing by construction a sense-annotated corpus, admittedly noisy but with large coverage. The method will be guided by pre-existing lexicons (in particular Wiktionary, available for many languages), and will make the best use of pre-trained transformer-based language models. The project also includes a part on the generation of definitions of these induced senses, as well as their use in a neural machine reading comprehension system, in order to improve its performance and the interpretability of its decisions.


  • The topic of this PhD position is more specifically the semi-supervised sense and frame induction part, using Wiktionary senses as constrained clustering seeds, and the grouping and structuring of induced senses into "semantic frames". The latter involves grouping occurrences of predicative lemmas, based on similarities of their argument structures observed in corpora, and grouping their semantic arguments into induced semantic roles.


Fairness by Explanations in Algorithmic Decision Making Full funded PhD, CNRS, LAMSADE, IRISSO Université Paris Dauphine, PSL



  • We offer a PhD contract for three years, funded by the CNRS Interdisciplinary Mission for brilliant students holding a MSc in Computer Science (or near disci- plines) open minded to the social sciences and to interdisciplinary challenges (a degree or a course in social sciences will be appropriately appreciated).


  • The topic concerns the study of autonomous artifacts design with enhanced deci- sion autonomy, satisfying some notion of “fairness”.


  • However, instead of consid- ering fairness as an “objective” property, related to a ressource allocation norm, we prefer study this notion as a subjectively defined necessity by one or more stakeholders of the decision process to which the autonomous artifact is related.


  • Under such a perspective we consider as component of such types of fairness the satisfaction of explicability, interpretability and accountability requirements.


  • The PhD student will be enroled with the Doctoral School of Université Paris Dauphine, PSL, and the research will be carried within the LAMSADE (https: //www.lamsade.dauphine.fr; the Computer Science Laboratory of the University) in cooperation with IRISSO (https://www.irisso.dauphine. fr; the Social Sciences Laboratory of the University). Motivated students can inquire or apply writing to Alexis Tsoukiàs ([email protected]).


  • Ap- plications should include a full CV, an academic record and a short research state- ment about the topic.


Doctoral thesis offer at GREYC - University of Caen Normandy



  • DOCTORAL THESIS "Effective declarative approaches for extracting patterns from closed intervals"


  • In data mining, many methods take as input a binary context for extracting patterns where each instance is described by binary attributes (or descriptors).


  • This poses a limitation when it comes to processing digital databases where a discretization step is required to find a binary representation that keeps the information from the original data. The interval structure patterns field is a formal framework for extracting closed interval patterns from numeric databases.


  • We are interested in the extraction of closed interval patterns with effective declarative approaches based on Constraint Programming and/or Linear Programming. These approaches have the ability to adapt to different data mining tasks without changing the resolution algorithm. Our goal is to offer efficient PPC models without discretization steps to efficiently extract patterns in digital databases .


  • We then use these models for the extraction of patterns of intervals of interest to the user via a learning step. We envisage the study of the method that will be proposed on the extraction of interval patterns in molecular databases in order to characterize the common properties related to their activity.


  • Keywords: Interval Pattern Mining, Constraint Programming, Structure Pattern, User Preferences, Learning.


  • *Greyc :* laboratory, CODAG team, University of Caen Normandy.


  • *Location :* University of Caen Normandy.


  • *Duration:* 3 years.


  • *Funding :* ANR HAISCoDe Project.


  • Candidate profile :* Master in Computer Science or equivalent for engineering schools , introduction to research in the form of follow-up teaching , project or internship. Solid knowledge of constraint programming, linear programming and pattern mining is appreciated.


  • To apply, please send your application (CV, cover letter , Bachelor's and Master's transcripts) to the following addresses: [email protected], [email protected] and [email protected].


  • You can find more details in the attached file.


  • *Deadline for full file: 28/04/2022*


PhD. and Post-doc Positions in AI Planning at Grenoble Alpes University



  • The University of Grenoble Alpes welcomes applications for one PhD. position and one Post-doc position (12 months possibly renewable) in the area of AI Planning.


  • Do not hesitate to contact me for more details.


  • Best regards, Damien Pellier


  • Université Grenoble Alpes Laboratoire d'Informatique de GrenobleBâtiment IMAG — 700 avenue CentraleCS 40700 - 38058 Grenoble cedex 9Bureau


  • : 365 / Tel: (+33) 4 57 42 15 39 http://membres-liglab.imag.fr/ pellier


PhD position - Response generation models for solving multi-faceted information needs



  • PhD proposal in Computer Science, main research fields : information science, information retrieval and access, natural language processing Keywords: language generation, data-to-text, machine learning, deep learning This position is proposed in collaboration between the ISIR laboratory (Sorbonne Université, Paris), the IRIT laboratory (Toulouse) and the Ecovadis company with a CIFRE contact.


  • Context


  • The perspective of new information retrieval (IR) systems (e.g., search-oriented conversational systems or systems supporting complex search tasks) has fostered research on theoretical IR models either leveraging or supporting users’ interactions, for instance, through question clarification or interactive ranking models. However, very few works focus on the way of interacting with the user in natural language, which is critical for instance for conversational systems.


  • PHD objectives


  • The main objective of the thesis is to design question-answering models aiming at solving multi-faceted information needs. More particularly, given a document collection, our goal is to generate structured and complete answers, covering all facets of a complex information need.


  • To do so, approaches and models from information retrieval (IR) and natural language processing (NLP) will be necessary. Both research fields exploit Deep Learning (DL) techniques to model semantics underlying texts and generate new knowledge. More precisely, we showed in a premise work [DGS+22] the potential of data-to-text approaches [PDL19a, RSSG20, PDL19b] for complex answer generation.


  • Our long-term objective is to fit with the conversational search setting and to deal with users’ interactions / conversational context [EPBG19, TY20] as well as include search task-oriented features in the generation process [FWZ+20, ZZW+20]. Two main lines of research stand out:


  • - one is linked to the multiplicity of data sources (text, tables, figures, etc.) used to generate the output text and structure.


  • - another one is more linked to the user satisfaction regarding the output in itself.


  • The generated document should both be complete, understandable and explainable.


  • Application to industrial use cases will be envisioned in collaboration with the development team at Ecovadis. All our models will be evaluated on academic benchmarks, enabling quantitative evaluation and the publication of the obtained results.


  • Scientific supervisors:


  • The research work will be supervised by : ● Laure Soulier, Associate professor (ISIR laboratory, Sorbonne Université) - [email protected]


  • ● Karen Pinel-Sauvagnat, Associate professor (IRIT laboratory, Université Toulouse 3 - [email protected]


  • ● Lynda Tamine, full professor (IRIT laboratory, Université Toulouse 3) - - [email protected]


  • ● Sophia Katrenko, PhD, Ecovadis - [email protected]


  • Location : The doctoral student may either be based in ISIR in Paris or in IRIT in Toulouse (preferably in Toulouse, but might be discussed with candidates). Regular exchanges will take place by video-conference as well as physical meetings.


PhD and postdoc positions on Intelligent handling of imperfect data (LaBRI, Bordeaux)



  • Dear all, The LaBRI research lab at the University of Bordeaux is currently seeking highly motivated candidates with experience in knowledge representation and reasoning, databases, and/or logic in CS to take part in the INTENDED AI Chair (intended.labri.fr), whose aim is to develop principled and effective methods for handling imperfect data.


  • Currently, one PhD position and one 2-year postdoc position are available. The start date for the PhD position is October 1, 2022. There is more flexibility for the postdoc position, but ideally it would start before the end of 2022.


  • Details on the positions, research environment, and how to apply, can be found on the project website:


  • https://intended.labri.fr/hiring.html


  • The first batch of applications will be reviewed in early May, so for fullest consideration, applications should be submitted by *May 1, 2022*.


PhD position - Explainability and interpretability in probabilistic planning



  • We propose a thesis offer within the LARSEN team of LORIA and INRIA in Nancy.


  • This thesis will focus on questions of explainability and interpretability of solution strategies in probabilistic planning , for example in the context of human-robot collaborations.


  • The thesis will start in September or October 2022 for a period of 3 years.


  • Application deadline: 2 May 2022.


  • Details of the offer and procedure to follow: https://recrutement.inria.fr/p ublic/classic/en/offers/2022-04720


Engineer / Postdoc: Deep Learning-Driver Risk prediction from telematics data



  • Duration : 1 year


  • Début : As soon as possible


  • Salary : 55 000 €


  • Institution : Institut Polytechnique de Paris / Telecom SudParis Palaiseau, France


  • Supervisor : Prof. Mounîm A. El Yacoubi Professor, HDR


  • Institut Polytechnique de Paris / Telecom SudParis / Institut Mines-Telecom SAMOVAR, CNRS, UMR 5157


  • Address: 19 place Marguerite Perey 91120 Palaiseau, France Phone: (+33) 1 75 31 44 53 (+33) 6 83 70 12 89 email: [email protected]


  • Candidate: - Master 2 or engineer, or Postdoc, in Data Science, Artificial Intelligence, Signal and Image Processing, etc.


  • - Solid experience in Machine Learning, in particular Deep Learning, supervised and unsupervised techniques


  • - Excellent analytical skills for solving real problems


  • - Excellent programming skills (Python)


  • - High level publications