How are experts identified on ideXlab?

The sources indexed by ideXlab are websites of major publishers, databases of scientific articles in "Open Access", patent databases, company databases. This architecture allows us to access in real time a very large number of documents that are always up to date, and to quickly integrate new data sources when they appear or if a customer asks us to. What these sources have in common is that they offer documents that have been read and validated by peers (peer-reviewed) or proofreaders, which gives a high degree of reliability to the results obtained following a query. 

Our search engine allows to identify and download all documents (scientific articles and patents) that contain the keywords used in a query: on ideXlab, experts are thus identified thanks to the different documents they have published. 

Once these operations are performed, our ranking algorithms will perform a textual analysis of all the downloaded documents and their authors in order to rank the experts in decreasing order of relevance (see article "How are the experts ranked on ideXlab?" for more details).

Globally, this method allows to rank the results not around the documents but around the most relevant experts. Indeed, ranking the results around the most cited documents (as in the case of Google Scholar for example), has the disadvantage of privileging possibly old documents that have had 'time' to be cited. Ranking around the most relevant authors, on the other hand, makes it possible to find the most recent papers of the most relevant authors, even if they are not yet widely cited. The other advantage of proceeding in this way is to immediately identify the authors with whom one could start a conversation or establish a partnership: one does not waste time contacting the author of a highly cited paper who has only occasionally worked on a subject.