Post-doctoral student in Data Sciences for CEFIPRA French-Indian project (H/F)

Key info

Organisation

Institut Mines Telecom- Télécom School of Management

Location

Evry, Ile De France, France (45 minutes from Paris center)

Industries

Big Data Algorithms

Position type

Full time contract-12 months

Salary

According to experience

 

Post-doctoral student in Data Sciences for CEFIPRA French-Indian project

 

Project description

 

CEFIPRA is a french-indian research project on modelling customer behavior on travel marketplaces for e-commerce.

Objectives of the project
The project is based on data from a web marketplace connecting local travel agencies specialized in road-trips and motorized travels, with passionate travelers ; Planet Ride. The company handpicks these agencies based on their quality of service to bring them extra visibility and additional business. Basically the marketplace delivers leads that agencies would exploit to sell road-trips and travels. Those leads come with data expressed by the client on the marketplace website, but no categorization or sorting are made between those lead even though lots of data is gathered during request process : navigation history, request history, geographic position, etc. Categorization in term of social context, potential expectation, expected time before time or even probability of booking would be a real advantage for the travel agencies during their lead processing. Cefipra project aims at building models and tools that could reveal buyers profiles or any patterns to help the decisions of travel agencies. These models and tools should improve the lead conversion of agencies and reinforce the insert of the platform for agencies.
Two main problems will be addressed during this project, mixing customer behaviour modelling, data analysis, preference modelling and machine learning. Customer behaviour analysis and modelling should provide useful inputs for the data analysis part, which in turn will lead to the tuning of machine learning algorithms modelling the preferences of the customers, and thus providing recommendations for the travel agencies. The resulting models and algorithms will be integrated by Planet Ride and Pandora in the marketplace.

Missions

State of the art of the literature on user experience on web platforms and specifically user behaviors on travel.

  • Descriptive analysis of the data obtained from Planet Ride platform, web server logs and other sources such as Google Analytics will be done.
  • Data analysis and modeling to segment travel marketplace users and to carry out predictive analysis of customer behavior and to segment travel marketplace users. Two main problems will be addressed during this project, mixing customer behaviour modelling, data analysis, preference modelling and machine learning. Customer behaviour analysis and modelling should provide useful inputs for the data analysis part, which in turn will lead to the tuning of machine learning algorithms modelling the preferences of the customers, and thus providing recommendations for the travel agencies. The resulting models and algorithms will be integrated by Planet Ride and Pandora in the marketplace.

Skills required

  • Ph D in Econometry or computer Science, or Graduated from a Scientific or Engineering School or a Master’s program in Machine Learning, Statistics, Mathematics or Applied Mathematics.
  • Interest in e-commerce and new technologies
  • Experience in using Machine Learning methods and in statistical computing and modelling.
  • Experience with a statistical tool adapted to big data analysis, such as Python, R or Matlab.
  • Knowledge of SQL or Java and Computer Science skills is a plus.
  • Strong mathematical skills.
  • Very good English skills.

Job location

Evry

Application

Send us your CV, with the reference “Cefipra Data Scientist “

Contact:  Christine.balague@telecom-em.eu