Section: New Results

Computational Social Sciences

Computational Social Sciences (CSS) is making significant progress in the study of social and economic phenomena thank to the combination of social science theories and new insight from data science. But while the simultaneous advent of massive data and massive computational power has opened exciting new avenues, it has also raised new questions and challenges.

Almost ten years after the first enthusiasms for "big data" in social science, P. Tubaro has undertaken a reflective effort to look back at progress made so far and at directions for the near future. She edited a special issue of Revue Française de Sociologie on the effects of data both on society itself and on the scientific disciplines that engage with it [46], of which she co-authored the introduction [13].

Meanwhile, four data-based studies are being conducted in TAU, about labor (hiring, working on Internet, quality of life and economic performance), about nutrition (health, food, and socio-demographic issues), around Cartolabe, a platform for scientific information system and visual querying and around GAMA, a multi-agent based simulation platform.

Labor Studies

Participants: Philippe Caillou, Isabelle Guyon, Michèle Sebag, Paola Tubaro

Post-docs; PhDs: Olivier Goudet; François Gonard, Diviyan Kalainathan, Thomas Schmitt

Collaboration: Jean-Pierre Nadal (EHESS); Marco Cuturi, Bruno Crépon (ENSAE); Antonio Casilli (Telecom); Thierry Weil (Mines); Jean-Luc Bazet (RITM)

A first area of activity of TAU in Computational Social Sciences is the study of labor, from the functioning of the job market, to the rise of new, atypical forms of work in the networked society of internet platforms, and the quality of life at work.

Job markets Our first study in the domain of job markets (Th. Schmitt's and F. Gonard's PhDs [12], [11]) tackled the matching of job ads and CVs. This study, funded by the Lidex Institut de la Société Numérique (ISN) at Univ. Paris-Saclay, was conducted in collaboration with EHESS, on data provided by the hiring Web agency Qapa (for blue-collars and temporary jobs) and by Association Bernard Gregory (for scientists in industry). Among other difficulties, this study revealed that for both qualified and unqualified job sectors, job seekers and recruiters do not speak the same language [143]. This first study will be continued and extended along two directions: counterfactual analysis (What would be my options if I had this additional skill ? DataIA project Vadore, coll. ENSAE and Pôle Emploi), and the recommendation of vocational training (BPI-PIA contract JobAgile, coll. EHESS and Qapa). Both projects start end 2018.

The platform economy and digital labor Another topic concerns the digital economy and the transformations of labor that accompany the current developments of AI. P. Tubaro has researched the so-called "sharing economy" and ideals of social change associated to the economic model of the platform [33]. However, the platform economy is also disrupting traditional industries. CNRS's MITI office has funded a research on the effects of online services for the restaurant sector (such as La Fourchette, Trip Advisor, Yelp) on working conditions and quality of service. This project involves P. Tubaro, P. Caillou and partners at Telecom ParisTech and Paris Dauphine University.

Ongoing research is exploring online platform labor and its linkages to the development of AI. In collaboration with A.A. Casilli (Telecom ParisTech), P. Tubaro has received funding to conduct research on this topic from the Union Force Ouvrière (OPLa project), from France Stratégie (a Prime Minister's service), and from MSH Paris-Saclay (DiPLab project). A recent grant from DARES (French Ministry of Labor) will enable exploring labor changes in B2B platforms (with O. Chagny of IRES, a unions-funded think-tank).

Quality of life at work. A study, funded by ISN, examined the relationship between the quality of life at work (QLW), and the economic performance of companies [113]. The management and economics literature has already established a correlation between QLW and economic performance [76]. The question that we are currently addressing regards the direction of causality: do profitable companies pay more attention to the QLW ? Or do companies paying attention to QLW tend to be more profitable ? This project (coll. RITM Univ. Paris-Sud, SES Telecom ParisTech, Ecole des Mines, La Fabrique de l'Industrie) combines data at the individual level (DARES, Ministère du Travail) and at the company level (Secafi); cutting-edge causality algorithms are applied to address the question, and handle confounder variables such as the sector of activity.

Health, food, and socio-demographic issues

Participants: Philippe Caillou, Michèle Sebag, Paola Tubaro

Post-docs; PhDs: Nayat Sanchez-Pi

Collaboration: Louis-Georges Soler, Olivier Allais (INRA)

Another area of activity concerns the relationships between eating practices, socio-demographic features and health.

The Nutriperso project (IRS Univ. Paris-Saclay, coll. INRA, CEA, CNRS, INSERM, Telecom ParisTech and Univ. Paris-Sud) aims to: i) determine the impact of food items on health (e.g., related to T2 diabetes); ii) identify alternative food items, admissible in terms of taste and budget, and better in terms of health; iii) emit personalized food recommendations (noting that general recommendations such as Eat 5 fruit and vegetable per day are hardly effective on the targeted populations. Based on the Kantar database, reporting the food habits of 20,000 households over 20 years, our challenge is to analyze the food purchases at an unprecedented fine-grained scale (at the barcode level), and to investigate the relationship between diets, socio-demographic features, and body mass index (BMI). The challenge also regards the direction of causality; while some diets are strongly correlated to high BMI, the question is to determine whether, e.g., sugar-free sodas are a cause, or a consequence of obesity, or both.

Previous research in this area included the study of eating disorders and their relationship to people's social network and usages of technology [18].

Scientific Information System and Visual Querying

Participants: Philippe Caillou, Michèle Sebag

Engineer: Anne-Catherine Letournel, Jonas Renault

Collaboration: Jean-Daniel Fekete (AVIZ, Inria Saclay)

A third area of activity concerns the 2D visualisation and querying of the scientific expertise in an institute/university, based on their scientific production, given as a set of articles (authors, title, abstract). The Cartolabe project started as an Inria ADT (coll. Tao and AVIZ, 2015-2017). It received a grant from CNRS (coll. Tau , AVIZ and HCC-LRI, 2018-2019). Further extension proposals, in collaboration with the department of bibliometry from Univ. Paris-Saclay, are under submission at the time of writing.

This project was initially devised as an open-source platform, aimed to answer burning questions, as the growth of academic organization prevents anyone from having a precise knowledge of who does what in the organization: Who is expert in a topic (described as a bag of words)? How are topics related? What are the rising topics? (see also Section 6.3)

Its development and the interaction with the beta-user scientists using it, increasingly raises new questions at the crossroad of human-centered computing, data visualization and machine learning: How to deal with poly-thematic researchers? How to take advantage of the fact that researchers have ideas about their relevant scientific neighborhood, and learn person-dependent metric?

Multi-Agent based simulation framework for social science

Participants: Philippe Caillou

Collaboration: Patrick Taillandier (INRA), Alexis Drogoul and Nicolas Marilleau (IRD), Arnaud Grignard (MediaLab, MIT), Benoit Gaudou (Université Toulouse 1)

Since 2008, P. Caillou contributes to the development of the GAMA platform, a multi-agent based simulation framework. Its evolution is driven by the research projects using it, which makes it very well suited for social sciences studies and simulations.

The 1.8 version of the platform[20] brings new capabilities required for social science research, such as High Performance Computing to explore the simulation, Co-Modeling to link projects, advanced agent architectures to model complex behaviors and advanced visualization to display nice 3D representations for exploration and presentations.