LReM interview

Sinan Bekka, Collaborator of Project Manager of Grande Marche Europe of La République en Marche (French progressist political party launched in 2016 by Emmanuel Macron), share with us his vision of the central role of Semantic Analysis to understand Frenchs opinion.

Present your position and your main missions at La République en Marche.

I am in the European Pole of La République en Marche. Currently I’m working essentially on the European campaign but during the Grande Marche for Europe I was helping with national coordination and also analysing and sharing the results.

Why did you put in place a Semantic Analysis solution for your verbatims?

Our method always starts from what the French people tell us in the field. The objective of Semantic Analysis is to go beyong closed-questions in surveys and to let the French people express themselves sincerely and freely on their feelings about Europe.

Why did you choose Proxem for your projects? Could you describe your feeling about the implementation?

It is not the first time that we have consulted the French people in this way. Our political movement was first born from the Grande Marche (Big Movement) in 2016. At this time we already used Proxem. Since then, the recurrence of our consultations (mainly online) had pushed us to have the analysis done directly by our teams, in an autonomous way. Therefore, the ease of use of the Proxem Studio helped us greatly, as well as the training and coaching sessions by the Proxem teams.

We encountered some obstacles, including to know how to manage words that had a similar spelling (for example “marche” and “marché” in French). We don’t have language training and have sometimes needed the help of Proxem’s expert teams to solve these types of linguistic challenges.

The main difficulty was the bias of the analysis. By analyzing the answers ourselves and already having an idea of ​​what we wanted to show, it was sometimes difficult to take a step back. Proxem’s teams helped us to refocus and complete some things that we missed.

Could you describe the results of our solutions?

The analysis allowed us to built a very solid analysis that we presented the 26th of September, 2018, in front of more than 900 people.

For example, the theme that emerged most from surveys was the environment, which we had not necessarily anticipated.

We continue to consult citizens on a wide variety of topics. We look forward to the new feature that automatically detecs the different meanings of the same word to help us to manage the ambiguity of language*.

*For example, the meaning of the term “proximity” varies according to whether we speak of “proximity to law enforcement”, “closeness of policies” (with citizens), “near Paris”, “closeness of election” or “convenience stores”; similarly, this last expression has the synonyms “AMAP”, “short circuits”… It is the possibility to take into account all of the homonyms and synonyms which makes it possible to have a qualitative semantic analysis.

Discover all our interviews: here.