Post-doc/Research engineer position – Information Retrieval in MBSE (Model-Based Systems Engineering)

An Institute for Technological Research (IRT) is an interdisciplinary thematic institute that develops economic sectors related to its field through a balanced strategic public-private partnership. For this, it manages research programs coupled with technology platforms, conducts research and development projects at the international level, contributes to the engineering of initial and continuous trainings (qualifying professional training and/or degree delivering); and ensures the exploitation of the obtained results.
Part of its international development strategy, the Technology Research Institute IRT SystemX, is based in Paris-Saclay campus (France). One of the strategic domains addressed by the institute concerns industrial engineering activities. Engineering activities, and R&D activities in general, are increasingly reliant upon digital technologies in order to make manufacturers, whatever their size, more agile, more productive and more competitive in these activities (competitiveness costs and innovation). In this domain, SystemX proposes developing methods, processes and collaborative engineering software tools for complex systems, in the context of the extended enterprise, by making use of the potential of digital technologies.

ISC project
Within SystemX, you will be part of the ISC project: Collaborative System Engineering for complex systems.
The ISC project is part of the digital transformation of the management, definition and development activities of large systems. Each industrial or state actor have for several years been engaged in transforming their internal practices. For each partner, setting up modeling practices and capacities for analysis and digital simulation opens the way to new collaborative capabilities for building of solutions: joint exploration of architecture options, flexible impact analysis and evaluation of industrial or operational feasibility, thorough verifications of inter-operability, guaranteed consistency of definition increments etc.
The ISC project is divided into three parts:
i) Process, methods and tools towards collaborative systems engineering.
ii) Evaluation and exploration of architectures.
iii) Heterogeneous systems artifact consistency.

Position

Your mission will concern “Heterogeneous systems artifact consistency” part.

Technical context:
The development of complex systems involves the collaboration of many stakeholders. Industrial practices are evolving from « classical » to model-based engineering approaches (MBSE). In order to design the system, they produce many artefacts i.e., requirements and models that are correlated with each other’s, in a more or less explicit way. Many organizations maintain textual descriptions alongside graphical models, e.g. requirements, referential.
To support organizations in keeping their design artefacts consistent, semantic overlaps must be identified, and traceability links between artefacts make explicit. A critical step to reach this goal is to provide automation support to do so. A direction followed by a PhD Thesis study in the project is to bring techniques from the Information Retrieval (IR) domain into the MBSE. This study as primarily been conducted on textual engineering data (requirements), and could be extended to models.

About the study:
The study will focus on the information extraction from models and transformation of these models into textual descriptions to feed identified IR techniques. Due to the conciseness of models (short labels, lack of a proper sentences structure), and the implicit contextual knowledge, this stage is non-trivial task and may require external domain knowledge. The transformation of these models to text in natural or controlled language should preserve the semantics of the models and be adapted to the IR Techniques (i.e. improve retrieval effectiveness: standard recall and precision measures).`
The main mission will be to:

  • Perform applied research (theoretical and empirical studies),
  • Conduct experiments and build prototypes/demos,
  • Write research reports,
  • Collaborate with researchers, engineers and PhD Student to provide expertise on the subject
  • Transfer the research outcomes to the industrial project partners.
  • Report in conferences or in papers on research outcomes

The Post Doc / research engineer will report directly to ISC project manager. Level of responsability and Autonomy at work could be adapted according to his skills.
The position is based in Palaiseau until the end of  ISC project in March 2020.

Profile

Required qualifications:

  • /PhD in Computer Science, Computational Linguistics, natural language processing or related fields.
  • Expertise in one or more of the following domains: Information retrieval (IR) techniques – Algebraic models and Probabilistic models, Natural Language Processing (semantics oriented) and Natural Language Generation (NLG) – information extraction, Knowledge Base Management.
  • Strong analytical and problem-solving ability.

Great to have:

  • Good understanding of linguistics and language as a phenomenon.
  • Experience with non-English (French) NLP.
  • Experience with Deep Learning and Word Embedding.
  • Knowledge in Model-Based approaches (MBSE, MDE)
  • Published work in academic conferences/journals or industry circles.
  • Familiarity with Java, Python, or other programming languages, and general software development skills (source code management, debugging, testing, deployment, etc.)

Professional qualifications :

  • Independent worker with the ability to effectively operate with flexibility.
  • Team player with exceptional interpersonal and solution-oriented attitude.
  • Ability to work as a project: reporting on progress and communicating results.
  • Ability to communicate both speaking and writing (English, French is a plus).

 

Reference: DOP 097
Applications: jobs@irt-systemx.fr