POSTDOCTORAL FELLOWSHIP 
OPEN POSITION
UNIVERSITY OF LORRAINE
 
TITLE Online Hate Speech Against Migrants
 
COORDINATION Research coordinator: Crem, University of Lorraine (Pr. Angeliki Monnier) Co-coordinator: Loria, University of Lorraine (Dr. Irina Illina MCF HDR, Dr. Dominique Fohr CNRS)
 
TERMS AND VENUE This one-year position will be based at Crem (Metz/Nancy) and Loria (Nancy), France (University of Lorraine). The target start date for the position is April 1st, 2019, with some flexibility on the exact start date.
 
SELECTION CRITERIA PhD in Information and Communication Sciences, or Language Science. Research experience in social media and language data analysis. Very good mastery of the French language. Very good written and oral expression skills, in French and in English. Curiosity, open-mindedness. Autonomy, teamwork skills.
 
HOW TO APPLY  Applicants are requested to submit the following materials: • A cover letter applying for the position • Full CV and list of publications • Statement of Research (summary of research achievements and perspectives for upcoming researches) • Academic transcripts (diplomas, courses statements, grades, etc.) (unofficial versions are fine) • Recommendation letters are not obligatory but are strongly recommended.
 
Deadline for application is December 10th, 2018.  Applications are only accepted through email. All documents must be sent to angeliki.monnier@univ-lorraine.fr
 
Applicants will be interviewed by an Ad Hoc Commission on January 15th, 2019.
 
DETAILED DESCRIPTION OF THE RESEARCH PROJECT
This position is open as part of the Open Language and Knowledge project for Citizens (OLKi), carried out within the IMPACT/Lorraine University of Excellence (LUE) framework. Under the direction of 

 
 
 
 
2
 
Pr. Angeliki Monnier (Crem), Dr. Irina Illina and Dr. Dominique Fohr (Loria), the post-doctoral fellow will work on online hate speech against migrants.
Social context
According to the 2017 International Migration Report, the number of migrants worldwide has increased rapidly in recent years. This development is causing great public concern around the world, particularly in Europe. The economic crisis affecting some countries of the Old Continent also feeds feelings of insecurity, encouraging the development of anti-immigrant movements. The media are often pointed out for their tendency to depict refugees and migrants negatively, consolidating fears. A recent EU project has revealed a significant increase in hate speech against immigrants and minorities, who are often accused of being the cause of current economic and social problems. Participatory web and social media seem to amplify the intensity and scope of hate speech. The fight against racism and hatred on the Internet is currently one of the priorities of the French government. On September 20th, 2018, a report commissioned on this topic was given to the Prime Minister, containing twenty proposals to combat hate on the Internet. 
Scientific context and scope
The objective of this postdoctoral contract will be to study the context of the appearance of hateful contents (circumstances of emergence, locutors, dissemination processes, etc.), and to analyze the latter as linguistic productions (narrative approaches, speech acts, enunciation, etc.) in the light of the creation of a lexicon of hate speech in French.
Crem and Loria are already involved in this project. Owing to the technology of “neural networks” (deep learning), their collaboration aims to collect and shape a corpus of hateful expressions against migrants, but also to develop an app that could automatically detect hate speech in comments posted on the Internet, especially in media websites.
The objective of the Crem-Loria collaboration in this postdoctoral contract is to refine these initial results through qualitative analyses of online hate speech against migrants. The aim is to achieve a better understanding of the social phenomenon of hatred, as well as to improve the development of algorithms used to qualify language.
For this reason, the collaboration between Crem and Loria will take an iterative form, between the qualitative analysis of restricted corpora and the work with algorithms. It will cover both the constitution of the corpus (online data collection, search for expressions of hate) and the analysis of this corpus (creation and organization of the lexicon).
The research will focus on user-generated content (social networks, comments on media websites, etc.). Part of the analyzed data will come from the OLKi platform, which will also serve as a support for the evaluation of the algorithms developed.
 
FELLOW’S MISSIONS • Collect online data (hate speech against migrants), using technical solutions proposed by Loria. • Analyze these discourses using Humanities and Social Sciences (SHS) approaches: sociopragmatic contexts of hateful comments, linguistic analysis of speeches (speech acts, narratives, enunciation, etc.). Depending on the candidate’s profile, image analysis can be included in the research project, as an additional component. • Contribute to the development of a lexicon in French about online hatred against migrants (supervised by Crem and Loria).

 
 
 
 
3
 
• Write scientific articles based on research results, in collaboration with the supervisors and cosigned with them, to be published in scientific journals and / or to be presented at national and / or international conferences. • Attend regular meetings between the two teams. • Provide regular reports on the project’s progress and a final report at the end of it. • The fellow is expected to regularly participate in the Crem and Loria seminars and other research activities.
 
AFFILIATION UNITS
Crem, Center for Research on Mediations, comprises more than 230 researchers: approximately 80 tenured scholars, more than 90 doctoral students, 45 associate members and 7 staff members. Its researchers belong to 11 disciplines: nearly 90 % come from the Information and Communication Sciences, Language sciences, French, Literature and Art Sciences; about 10 % are specialists in English and Anglo-Saxon languages, Arabic, Germanic and Romance languages, Anthropology, Psychology and Sociology.
Pixel is one of the four Crem teams, with more than 40 members, including 16 tenured scholars, with a specialization in the field of the usages of information and communication technologies. Pixel researchers implement different methodological approaches: surveys, content analyzes, sociotechnical analyzes, usage observations, socio-historical analysis, etc. For years, Pixel has been developing a sustained research activity around several thematic areas: digital educational practices (online learning platforms, serious games), access to online information (search engines, information websites, social networks, micro-blogging, information monitoring), online collaboration (participatory work environments, watch and curation tools, viral dissemination of journalistic content) and creative industries (digital games, expressive games, video games). Website: https://crem.univ-lorraine.fr/lunite/equipe-pixel/ 
Loria, Laboratory of Research in Computer Science and its Applications is a joint research unit (UMR 7503), common to several institutions: the CNRS, the University of Lorraine and Inria. Since its creation in 1997, Loria's mission is to enhance and promote fundamental and applied research in Computer Sciences. The scientific work is carried out in 28 teams structured in 5 departments, 15 of which are shared with Inria, representing a total of more than 400 persons. Loria is one of the largest laboratories in Lorraine.
Multispeech is one of the 28 Loria-INRIA Grand Est teams. It comprises 12 tenured scholars. The Multispeech research project focuses on speech processing, paying particular attention to multisource (source separation, robust speech recognition), multilingual (foreign language learning) and multimodal (audiovisual synthesis) aspects. Its research program is structured in 3 axes: - the explicit modeling of speech, which exploits its physical dimension; - the statistical modeling of speech, which relies on techniques of automatic learning such as the Bayesian models (HMM-GMM) and networks of deep neurons (DNN); - the uncertainties related to the high variability of the speech signal and the imperfection of the models. Website: http://team.inria.fr/multispeech/