Missions

* work on methodologies for socio-semantic networks analysis and opinion dynamics analysis.
* develop innovative methods to fight online disinformation.
* work with a team of engineers on the development of a real-time macroscope for social dynamics.

Activités

Some of the tasks that the hired person will perform include:
* Develop new multi-modal methods for social networks mining in several languages based on complex networks analysis and machine learning. In particular, we will extent previous methods using temporal graph network and notions of causality in graph dynamics.
* Collaborate on open source scientific software tools and data analysis pipelines.
* Interact with team members across partner institutions to collaborate on interdisciplinary research.
* contribute to the preparation and redaction of academic articles.

Compétences

- PhD in Computer Science, Mathematics, Physics or a related fields.
- You have experience in complex networks analysis, web-mining and/or text-mining.
- You have worked with Machine Learning and Graph algorithms.
- You have a strong experience with deep learning frameworks such as TensorFlow and PyTorch.
- You have proficiency in Python and other programming languages used for data processing and data science.
- You have excellent communication skills and you are able to articulate clearly the technical needs, set clear goals and work within an interdisciplinary setting, communicating with other scientists.
- Proven experience (scientific publications, GitHub page or equivalent work experience) in performing analyses using machine learning classifier algorithms and/or networks methods.
- Experience of statistical or other programming languages (preferably Python) to manipulate large-scale datasets with social network information.
- Fluency in English.
- Good relationship, knowing how to combine team spirit and autonomy.
* Advantageous:
- Experience working with Big Data frameworks such as Apache Spark (PySpark), Hadoop, and HDFS.

Contexte de travail

The Paris Île-de-France Complex Systems Institute (ISC-PIF) is an interdisciplinary unit of the National Center for Scientific Research (CNRS), one of the largest and most prestigious French research organizations.
Interface between disciplines, but also between research organizations and higher education, the ISC-PIF is a place dedicated to the development of innovative and interdisciplinary research on complex systems at the crossroad between modeling, high performance computing and big data. Located in Paris, it catalyzes, since 2005, the emergence of common and interdisciplinary practices, facilitating access to skills, work areas and pooled research resources.

In the spirit of the digital commons, ISC-PIF develops and provides to academia and citizens "social macroscopes", i.e. platforms aiming at the observation and understanding of collectives at large scale through the analysis of public digital traces (see for examples "Politoscope" - http://politoscope.org - and "Climatoscope" - http://iscpif.fr/climatoscope).

The hired person will have full access to the ISC-PIF's Multivac platform, one of the CNRS flasgship platform in the domain of Big Data analytics that provides access to high performance computing, longitudinal databases with billions of records, collaborative coding environments and real-time APIs. S|he will work under the respondability of David Chavalarias (PI of the project and Director of ISC-PIF), in close collaboration with Maziyar Panahi (project leader of Multivac) and the computational science team of ISC-PIF. The NODES project takes place within a larger team of a dozen of scholars working in the domain of computational social sciences, AI and text-analytics.



Within the European NODES project (start Feb. 2023), ISC-PIF will develop methodologies to analyze social dynamics in near real-time and will contribute, via its macroscopes, to the prototyping of an open European Narrative Observatory. The focus will be on Climate Change, Health and Politics with an emphasis on understanding and fighting online disinformation. This prototype will be based on network science, text-mining and machine learning for the methodologies.

The hired person will work within a interdisciplinary team of researchers and engineers.