Industrial Doctorate at the Paris metro company (RATP)
- Developed a zero-cost, real-time questionnaire system through social media bots to gauge passenger sentiment during service disruptions, culminating in a published study.
- Data-enginered a dataset of over 30 million RATP-centric tweets and analyzed it using advanced NLP methods like Spacy, CamemBERT, and large language models such as GPT-4.
- Executed a detailed distress analysis based on social media data and internal train data from the major Line 4 incident in June 2023, providing actionable insights for customer service follow-up and internal operational audits.
- Devised an innovative psychometric instrument to quantify and improve the clarity of communication during traffic disruptions, mapping passengers perception of 63 key messages over 9 dimensions.
- Initiated a strategic initiative to implement statistical models in forecasting return to normalcy time estimates, so as to double the precision of estimations during disruptions.
- Crafted key performance dashboards for the executive committee in charge of passenger experience, marketing and services.
Animation 1: The most prevalent emotion experienced during traffic disruptions is helplessness.
Video 1: Presentation of the research stakes in the summer of 2021 for the social networks of ENS Ulm.
Disorientation during the Covid-19 pandemic (RA-COVID19 v11)
- At the beginning of the Covid-19 pandemic, we posited that the pandemic was fundamentally disorienting in a manner akin to spatial disorientation. We aimed to document and investigate these non-spatial forms of disorientation. (Hypothesis paper published on Global Discourse)
- We first distributed a qualitative questionnaire and performed a phenomenological analysis of the reports. We identified six emerging themes: temporal rift; temporal vertigo; impoverished time; tunnel vision; spatial and social scaffolding of time; suspended time. We drew a key distinction between episodic and existential forms of temporal disorientation, and we argued that the Covid-19 crisis is best conceptualised as a period of suspended time. (Qualitative paper published on Phenomenology and the Cognitive Sciences)
- Based on this qualitative analysis, we designed and validated a psychometric instrument aimed at measuring temporal and social disorientation. These two non-spatial forms of disorientation were the most prevalent ones during the Covid-19 pandemic. (Validation paper published on PLOS One)
- We distributed this instrument to thousands, including students from higher education institutions in France and the general population on Twitter. We analyzed 3,306 responses and discovered that perceptions of a slow passage of time and skewed distance judgments were associated with significant disturbances in all examined domains. These included temporal disruptions—the ability to project oneself into the future; the sense of a rift between pre-pandemic and pandemic times; locating oneself in time; recalling the order of past events—as well as an overall sense of social disorientation, and trauma-specific disturbances. (Analysis paper published on Nature Scientific Reports)
- Our results have policy implications: (1) perceived passage of time is interesting to monitor during crisis management as it can serve as a proxy indicator for a wealth of subjective disruptions, (2) these subjective disruptions, in turn, were mostly driven by social disorientation during Covid-19. Public authorities should seek to provide clear future temporal landmarks to prevent disorientation, as well as ensuring those at risk are not socially secluded.
Animation 2: Experiencing a slow passage of time was an indicator of social and temporal disorientation during the Covid-19 pandemic.