Abstract: The techniques disclosed herein enable systems to leverage publisher profile and reputation to detect malware and block malicious actors. This is accomplished by obtaining a description of a software application that is pending release from the publisher. The description defines the nominal functionality of the software application such as expected program behaviors and device driver accesses. The system then catalogues the actual functionality of the software application by analyzing the program code implementing the software application. Accordingly, the generative model can detect inconsistencies between the actual functionality and the nominal functionality. In response, the generative model can initiate an interaction with the publisher to resolve the inconsistency. The publisher can respond with a justification of the inconsistency which is evaluated by the generative model in accordance with a reputation score associated with the publisher.

Abstract: Described herein are various technologies pertaining to providing information to a user regarding behavior of a potentially unwanted application. In response to this information, the user can determine action(s) to take regarding the potentially unwanted application. Further, optionally, based upon action(s) taken by the user, an adaptive component can modify information collected by a collector component.

Abstract: Described herein are various technologies pertaining to providing information to a user regarding behavior of a potentially unwanted application. In response to this information, the user can determine action(s) to take regarding the potentially unwanted application. Further, optionally, based upon action(s) taken by the user, an adaptive component can modify information collected by a collector component.

Abstract: One or more techniques and/or systems are provided for identifying a location interest inference for a first user. A first set of user signals (e.g., search history, social network posts, etc.) associated with the first user may be evaluated to identify a first user location interest pattern indicative of location interests of the first user. Social signals (e.g., phone calls, emails, photo tags, shared content, etc.) between the first user and other users may be evaluated to identify a second user having a social activity relevance score above a relationship threshold with respect to the first user. A second user location interest pattern may be generated for the second user based upon user signals associated with the second user. The first user location interest pattern and the second user location interest pattern may be aggregated to create the location interest inference indicative of refined locational interests of the first user.

Abstract: Online sellers may be ranked based on feedback given by people trusted by an individual user. The user may trust people in their social networks, as well as people who may be experts in a particular field, and the seller's ranking may be calculated by weighting reviews or feedback from trusted people higher than people unknown to the user. When used with a social campaign management system, ranking of products from multiple online sellers may include coupons or incentives that are available through the user's social network, as well as discounts or incentives that may be targeted to the user's status within their own social network.