The present invention relates to a method for providing information on a product based on high-speed similarity calculation, and a recommendation server therefor. The method comprises the steps of: allowing a data collection server to collect, from user terminals, a plurality of behavior logs each including a user identifier and a product, and transmit the collected behavior logs to an analysis server; allowing the analysis server to delete some of the plurality of behavior logs; allowing the analysis server to calculate similarity by performing similarity calculation on the plurality of behavior logs of which some are deleted; allowing the analysis server to generate information on a product according to a result of the similarity calculation; and allowing the analysis server to transmit the information on the product to an application server. Thus, the load of the system can be reduced.;COPYRIGHT KIPO 2020
展开▼