Rnest by La mètis

Rnest is a software offering a new web search experience for professionals. After the keyword approach (Yahoo 1994) and the popularity search (Google 1998), La mètis offers with Rnest a personalized web search tool to face the information chaos. Rnest automates the tasks of a user searching for information on the web: visiting preferred sites, selecting and reading articles, extracting knowledge, and finally creating a synthesis. While a human can read a few dozen articles from a dozen known websites, Rnest reads 250,000 articles from thousands of starting sources. Rnest extracts and makes available an average of 2,500 relevant results, from which it selects 50 must-see items, which are the priority. It also suggests new concepts, places, organizations, people and sources of information related to the field in question to enable a more in-depth search in an iterative approach. Rnest thus makes it possible to explore entire areas of the web that the user does not know about because of a lack of time. For each web search, Rnest performs the amount of work of 10 employees during a year. To do this, Rnest implements an "explainable" and traceable artificial intelligence, generating confidence in the origin and processing of web data.