Postdoctoral position: Geographical enrichment of Twitter data
Ref N° DL07-1718-OCRN
Deadline: Opened till position filled
Starting date: September 2017
University of Technology Troyes, UMR CNRS 6281 ICD (Institut Charles Delaunay) / Département « Recherche Opérationnelle, Statistiques Appliquées, Simulation » / Projet DataLab AnalyTrics (Analytics & Metrics)
The candidate will participate in the project OCRN with the BRGM which is the french public institution for Earth Science applications in the management of surface and subsurface resources and risks. This project OCRN deals with the management of natural disaster crisis management by gathering user generated informations such as social media.
In the case of social networks like Twitter it is difficult to identify the geographical location of the tweets. However, the implementation of a GIS remains strongly dependent on our ability to enrich the tweets collected by an inference of the geographical position of the messages.
Several algorithms in the literature propose to identify the geographical position of emission of a message either from its content, or by assessing the quality of its links with its contacts.
The successful candidate will work on the following tasks:
(1) conduct a state of art for the implementation of a multi-indicator approach for tweet geographical position identification,
(2) implement a geoinference online tool by integrating an iterative approach for geographic enrichment of a tweets database,
Particular attention will be paid to the privacy of the profiles whose tweets will be analyzed.
To qualify, applicants must hold a Ph.D in computer science or data mining. Knowledge in geographical Information system and Twitter API would be appreciated.
Grant: 1-year postdoctoral position with possible renewal upon results and mutual consent.
Salary: remuneration based on the postdoctoral salary scale at UTT
To apply send us your CV and a letter of motivation stating how your competencies can fit the research agenda. In the subject of your email put the reference DL07-1718-OCRN.
(c) GdR 720 ISIS - CNRS - 2011-2015.