Using clustering engines in document retrieval

An Internet search using conventional search engines often leads to sets of results which, although arranged in a certain order, do not highlight the main themes covered by the search. References are not presented in meaningful blocks. As a result, the user cannot directly access the various themes selected in the search. The mapping of the subject is therefore not available.

To provide an answer to this problem, some companies have made "clustering engines" available via the Internet which, based on a question posed by the user, will provide the results by dividing them into significant blocks with a precise heading for each block. This enables the user to see almost instantaneously the different themes that appear in the search result, and also to access each of the answers present in the blocks of information. The results are serialized using data analysis programs that are transparent to the user.