METHOD FOR PRIVATE KEYWORD SEARCH ON STREAMING DATA

At step (210), a buffer (35) is provided having each slot (40) therein initialized to an E(O). "Providing" the buffer (35) refers simply to the buffer (35) being available in the described condition, whether the current user, obtained from an independent source, created it or simply is available for use to carry out the methods described herein. Typically, the buffer (35) will be initialized in a secured. When there are no more documents, the query at (step 225) is answered in the negative, and so control passes to (step 230) at which time the buffer is sent to a secured environment for decryption. The process then ends at (step 240). Of course, a new data stream (505) may then be processed, and the process would begin all over at (step 200). The entire process, as described herein, may be performed by multiple public sources on multiple data streams, each source then transferring its buffer of data to the secured environment for decryption upon completion L'invention concerne un procédé de recherche de mots-clés privée sur des données en flux dont la recherche ne révèle pas les mots-clés concernés par celle-ci, si de tels mots-clés ont été trouvés, ni les documents dans le flux de données qui sont sauvegardés.