Sites Inria

Version française

Semantic Web


The Wimmics Team and Viseo Give Meaning to Data

Fabien Gandon, Inria (équipe Wimmics) - © Inria / Photo G. Maisonneuve

The Wimmics Inria–I3S project team and the Viseo R&D centre have created a joint laboratory (LabCom) to widen the scope of research on natural language processing and the Semantic Web. Dubbed SMILK, it will receive funding from the French National Research Agency over a period of three years. The goal is to apply research findings to the design of commercial tools for targeted marketing

All human activities generate data available through the Web. But the volume of this data—input by humans, generated by machines, or recorded by sensors—is constantly growing. The task now is interpreting this churning torrent of online data to better manage its content, make the right decisions, define the best strategies, and optimize risk management in various vertical markets and fields.


The SMILK LabCom: Bilateral Collaboration


SMILK wants to find meaning and links between data from social media so that companies can learn more about their customers—their preferences, habits, relations, and expectations—for market intelligence and targeted marketing purposes. They could then use this information to offer personalized products and services.


To accomplish this, the LabCom brings together Viseo expertise in natural language processing and Inria experience with Linked Open Data and the Semantic Web.

  • Such processing takes a given quantity of text and analyses it to yield structured information, including the subjects it addresses, the people or places it mentions, and the polarity of opinions it expresses.
  • The goal of open data is to link up huge online repositories and databases to pool data on a particular topic and perform other operations.


Enhancing and Extending Research


SMILK will try to find answers to the following questions:

  • How can we extract the semantic content of data from social media?
  • How can we structure this data and define its semantic relationships?
  • Can natural language processing help with this structuring and vice versa?
  • How can we navigate through this data?
  • What methods can we use to group data and use it to assign values to indicators?


This collaborative effort with the Wimmics team led by Fabien Gandon will allow Viseo to add to its automatic language processing expertise through Semantic Web insights and methods. This will help it optimize R&D costs and develop its range of business intelligence products. Viseo would like to focus on software that bolsters a company’s IS (usually an ERP) by interfacing with the Web.  Through this joint laboratory, Wimmics will gain access to new real situations and data, and understanding needs in the field, to enhance and extend its research.

Keywords: WIMMICS Language Semantic Web Viseo