Abstract: Code of a particular application is analyzed against a semantic model of a software development kit of a particular platform. The semantic model associates a plurality of application behaviors with respective application programming interface (API) calls of the particular platform. A set of behaviors of the particular application is identified based on the analysis of the code and a particular one of the set of behaviors is identified as an undesired behavior. The particular application can be automatically modified to remediate the undesired behavior. The particular application can be assigned to one of a plurality of device modes, and access to the particular application on a user device can be based on which of the plurality of device modes is active on the user device.

Abstract: A safety event is determined as affecting a user based at least in part context data collected at a user device associated with the user. In some aspects, context data is detected from sensors on the client device, the context data describing a present context of the user. A deviation of the present context from a historical context is determined to be beyond a threshold. Determining that the deviation is beyond the threshold can be determined to correspond to a safety event potentially jeopardizing safety of the user. In some aspects, an action can be launched in response to determining the safety event.

Abstract: A system for controlling accesses to network enabled devices includes a network interface over which a hub communicates with network enabled devices, a processor, and a multilayer access control layer. The access control layer includes instructions that, when executed by the processor, cause the processor to detect, at the hub, a request representing an attempt by an application executing on a remote host device to access a network enabled device communicatively coupled to the hub, characterize the request according to a user of the remote host device, the application making the attempt, and the network enabled device, and determine whether to allow or deny the request based upon the characterization and a plurality of rules. The rules may include definitions of access rights, with respect to the network enabled device, for users, applications, commands or queries made by applications, remote host devices, and network domains.

Abstract: Particular embodiments described herein provide for an electronic device that can be configured to receive a request to share data, determine metadata for the data to be shared, communicate the metadata to a social drive, where the social drive is separate from the electronic device and the data is not located on the social drive, and communicate the shared data to a member of the social drive when the member requests the data.

Abstract: To provide a more seamless experience across multiple devices, task streaming systems and methods allow a user to create “task-contexts” and manage metadata of files stored across multiple data storage devices and user preferences associated with capabilities of the multiple devices for operating on the file. Furthermore, the task streaming systems and methods are provided to allow task-contexts to be shared from one device to another device. A task-context specifies one or more files and one or more operations to be performed on the one or more files. By providing a task-context from one device to the other device, a user can accomplish a task with a particular file and seamlessly transition between devices with minimal disruption and effort.

Abstract: Distributed favorite files are stored in a favorite storage device. An agent dynamically monitors usage statistics for a plurality of files, wherein the plurality of files are located on a plurality of devices, designates a first file of the plurality of files as a favorite file based on usage statistics for the first file, and in response to designating the first file as a favorite file, stores a copy of the first file on a favorites storage device.

Abstract: Securely distributing a backup file includes identifying a file to be stored, generating a plurality of file chunks from the file, transmitting from a first device, each of the plurality of file chunks to one of a plurality of storage locations, and generating, at the first device, a map that identifies the storage location for each chunk. Retrieving a distributed backup file includes receiving a request to retrieve a file, obtaining a map corresponding to the file, wherein the map identifies a storage location for each of a plurality of file chunks of the file, retrieving the plurality of file chunks from the identified locations, and generating the requested file by combining the plurality of file chunks.

Abstract: The present disclosure relates to a system and method for performing antimalware scanning of data files that is data-centric rather than device-centric, In the example, a plurality of computing devices are connected via a network. An originating device creates or first receives data, and scans the data for malware, After scanning the data, the originating device creates and attaches to the data a metadata record including the results of the malware scan, The originating device may also scan the data for malware contextually-relevant to a second device.

Abstract: A safety event is determined as affecting a user based at least in part context data collected at a user device associated with the user. In some aspects, context data is detected from sensors on the client device, the context data describing a present context of the user. A deviation of the present context from a historical context is determined to be beyond a threshold. Determining that the deviation is beyond the threshold can be determined to correspond to a safety event potentially jeopardizing safety of the user. In some aspects, an action can be launched in response to determining the safety event.

Abstract: Code of a particular application is analyzed against a semantic model of a software development kit of a particular platform. The semantic model associates a plurality of application behaviors with respective application programming interface (API) calls of the particular platform. A set of behaviors of the particular application is identified based on the analysis of the code and a particular one of the set of behaviors is identified as an undesired behavior. The particular application can be automatically modified to remediate the undesired behavior. The particular application can be assigned to one of a plurality of device modes, and access to the particular application on a user device can be based on which of the plurality of device modes is active on the user device.