公开(公告)号：US11257100B2

公开(公告)日：2022-02-22

申请号：US16223340

申请日：2018-12-18

Applicant: SAP SE

Inventor： Iuliia Konstantinova ,  Ralph Debusmann ,  Michail Vasiltschenko ,  Anton Snitko ,  Günther Fischer

IPC: G06Q30/02 ,  G06N20/00 ,  G06F16/951 ,  G06N5/04

Abstract: Disclosed herein are system, method, and computer program product embodiments for crawling feedback sources and generating a product improvement recommendation. In an embodiment, a Product Optimization Crawler and Monitor (POCM) system may crawl feedback comments from different sources such as an Internet source or a customer feedback database. The POCM system may apply artificial intelligence, natural language processing, and constraint modeling techniques to the feedback comments to identify product features as well as a feedback category corresponding to the product feature. The feedback category may include a positive, negative, or neutral feedback category. Using this information, the POCM system may generate a summary of the feedback from different sources and/or generate product improvement recommendation. The product improvement recommendation may include suggesting that a component from a first product be replaced with a similar component from a second product.