| 摘要 |
The present disclosure describes one or more systems, methods, routines and/or techniques for automatic detection of fraudulent ratings and/or comments related to an application store. The present disclosure describes various ways to differentiate fraudulent submissions (e.g., ratings, comments, reviews, etc.) from legitimate submissions, e.g., submissions by real users of an application. These various ways may be used to generate intermediate signals that may indicate that a submission is fraudulent. One or more intermediate signals may be automatically combined or aggregated to generate a detection conclusion for a submission. Once a fraudulent submission is detected, the present disclosure describes various ways to proceed (e.g., either automatically or manually), for example, the fraudulent submission may be ignored, or a person or account associated with the fraudulent submission may be penalized. The various descriptions provided herein should be read broadly to encompass various other services that accept user ratings and/or comments. |
| 主权项 |
1. A method comprising:
hosting an application service that provides for users to enter submissions regarding a plurality of hosted applications; receiving, via the application service, a submission related to a targeted application of the plurality of hosted applications, wherein the submission expresses a review or amount of favorability of the targeted application; analyzing the submission to detect that the submission is undesirable, wherein the analysis includes aggregating a plurality of signals that indicate that the submission was not entered by a legitimate user of the targeted application, and wherein each of the plurality of signals is generated by considering the content of the submission and/or the relationship between the submission and other submissions received via the application service and/or information from the application service or a client device; and automatically imposing one or more penalties on the targeted application or user account(s) associated with the undesirable submission. |
