Know more

About cookies

What is a "cookie"?

A "cookie" is a piece of information, usually small and identified by a name, which may be sent to your browser by a website you are visiting. Your web browser will store it for a period of time, and send it back to the web server each time you log on again.

Different types of cookies are placed on the sites:

  • Cookies strictly necessary for the proper functioning of the site
  • Cookies deposited by third party sites to improve the interactivity of the site, to collect statistics

Learn more about cookies and how they work

The different types of cookies used on this site

Cookies strictly necessary for the site to function

These cookies allow the main services of the site to function optimally. You can technically block them using your browser settings but your experience on the site may be degraded.

Furthermore, you have the possibility of opposing the use of audience measurement tracers strictly necessary for the functioning and current administration of the website in the cookie management window accessible via the link located in the footer of the site.

Technical cookies

Name of the cookie


Shelf life

CAS and PHP session cookies

Login credentials, session security



Saving your cookie consent choices

12 months

Audience measurement cookies (AT Internet)

Name of the cookie


Shelf life


Trace the visitor's route in order to establish visit statistics.

13 months


Store the anonymous ID of the visitor who starts the first time he visits the site

13 months


Identify the numbers (unique identifiers of a site) seen by the visitor and store the visitor's identifiers.

13 months

About the AT Internet audience measurement tool :

AT Internet's audience measurement tool Analytics is deployed on this site in order to obtain information on visitors' navigation and to improve its use.

The French data protection authority (CNIL) has granted an exemption to AT Internet's Web Analytics cookie. This tool is thus exempt from the collection of the Internet user's consent with regard to the deposit of analytics cookies. However, you can refuse the deposit of these cookies via the cookie management panel.

Good to know:

  • The data collected are not cross-checked with other processing operations
  • The deposited cookie is only used to produce anonymous statistics
  • The cookie does not allow the user's navigation on other sites to be tracked.

Third party cookies to improve the interactivity of the site

This site relies on certain services provided by third parties which allow :

  • to offer interactive content;
  • improve usability and facilitate the sharing of content on social networks;
  • view videos and animated presentations directly on our website;
  • protect form entries from robots;
  • monitor the performance of the site.

These third parties will collect and use your browsing data for their own purposes.

How to accept or reject cookies

When you start browsing an eZpublish site, the appearance of the "cookies" banner allows you to accept or refuse all the cookies we use. This banner will be displayed as long as you have not made a choice, even if you are browsing on another page of the site.

You can change your choices at any time by clicking on the "Cookie Management" link.

You can manage these cookies in your browser. Here are the procedures to follow: Firefox; Chrome; Explorer; Safari; Opera

For more information about the cookies we use, you can contact INRAE's Data Protection Officer by email at or by post at :


24, chemin de Borde Rouge -Auzeville - CS52627 31326 Castanet Tolosan cedex - France

Last update: May 2021

Menu Logo Principal


Machine Learning process


Explicit articulation of uncertainty in science, especially in science involved in public policy, improves science, because it clarifies what future research is necessary, helps policy makers to evaluate scientific reports, and reminds the public about the limitations of current scientific knowledge.

We have built two complementary ontologies to measure the scientific uncertainty expressed in food safety documents. We have enlisted Machine Learning to help with three tasks:

  • To assist with coding complex documents such as food safety risk assessments
  • To assess the quality of the ontologies we devised (uncertainty and judgment ontologies)
  • To get insights into the causal processes of making uncertainty explicit. 


One straightforward strategy for the construction of an automated coding system for these documents is to use supervised learning techniques. Supervised learning aims at finding rules starting from a set of training instances. In our case, the training instances are the sentences (or small set of sentences) and their associated labels given by human coders. The goal is to extract rules that would allow a system to automatically label new sentences (or set of sentences) drawn from documents similar to the one used by the human coders. For instance, the system should be able to code a text input as ‘coded’ vs. ‘non coded’ and, if ‘coded’, as one of the categories present in the ontologies. Furthermore, if the learning system is trained from instances drawn from different contexts, differences in the learned rules could be enlightening about differences between these contexts. For instance, it could appear that the rules learned from American documents differ somewhat from the rules learned with documents from the European Union. Or that the rules change over time.

In the estimation, we used naïve Bayesian classifier, Support Vector Machines (SVM),kNearest Neighbor and Decision Tree algorithms. The best overall learners were Naïve Bayes and SVM.


We found that Machine Learning can do surprisingly well identifying complex meanings and probably can be helpful making suggestions to human coders. Our ML experiments show that our ontologies do enable a fairly consistent practice of coding.

Evaluating our ontology of judgment, we learned that elements of judgment are often communicated in relationship to one another. In our future work, we will try to exploit these relationships to identify judgment variables. Assessing our ontology of uncertainty, we found out that the deductive decision making process, that aids human coders, and that is reflected in the hierarchical structure of our ontology, makes the first large cuts fairly well and confusions tend to emerge between variables that our system defines as being closer.

And finally, we found some evidence that institutional factors do influence how predictable our ontology of uncertainty is.