Abstract: Telemetry data from a plurality of peer computers of a peer-to-peer network is aggregated via a computer network. Each of the plurality of peer computers sends telemetry data related to transfer of a digital content item within the peer-to-peer network. A content-acquisition request querying for a recommended content source to provide a first digital content item is received from a first peer computer of the peer-to-peer network via the computer network. A response to the content-acquisition request is sent to the first peer computer via the computer network. The response identifies a second peer computer of the peer-to-peer network that has the first digital content item as the recommended content source. The second peer computer is selected according to a peer selection metric derived from the telemetry data aggregated from the plurality of peer computers.

Abstract: A content-acquisition request is sent to a centralized management service computer via a computer network. The content-acquisition request may query the centralized management service computer for a recommended content source to provide a first digital content item. If a response to the content-acquisition request is received via the computer network and identifies a recommended peer computer of a peer-to-peer network as the recommended content source, a request to download the first digital content item is sent to the recommended peer computer via the computer network. If a response to the content-acquisition request is not received, a fallback request to download the first digital content item is automatically sent to a content-delivery network computer via the computer network.

Abstract: A plurality of sources storing portions of a digital content item that includes a plurality of pieces is identified. The plurality of sources includes one or more local storage machines of a computer and one or more peer computers of a peer-to-peer network. For each piece of the plurality of pieces of the digital content item, that piece is downloaded from a source. The source is selected from the plurality of sources according to one or more download metrics. The plurality of pieces is organized for installation on the computer as the digital content item.

Abstract: Embodiments for dynamically varying a buffering order of a digital content item are disclosed. One disclosed embodiment provides a computing device configured to receive content access information for a plurality of client devices, the content access information describing consumption of a digital content item provided according to a buffering order previously sent to each client device of the plurality of client devices. The computing device is further to dynamically update the buffering order based on the content access information to produce an updated buffering order and to send the updated buffering order to one or more client devices.

Abstract: Certain embodiments relate to use of aggressive peering requests, which enable a peer computing device to obtain desired digital content more quickly than typically possible in a P2P network. In certain embodiments, an aggressive peering request comprises a request that another peer computing device, to which the aggressive peering request is sent, dedicates substantially all of, or a disproportionately large amount of, its P2P resources to servicing a specific peer computing device. Other embodiments relate to identifying, based on accessed information, peer computing devices that are predicted to be available as an uninterrupted seed, and thus, can be used to increase download throughput in a P2P network.

Abstract: Telemetry data from a plurality of peer computers of a peer-to-peer network is aggregated via a computer network. Each of the plurality of peer computers sends telemetry data relating to transfer of a digital content item within the peer-to-peer network. A mitigation operation that modifies transfer of a digital content item between peer computers of the peer-to-peer network is performed according to one or more health metrics of the peer-to-peer network. The one or more health metrics are derived from the telemetry data aggregated from the plurality of peer computers.

Abstract: Embodiments disclosed herein generally relate to improving distribution of digital content in a peer-to-peer network. For example, future snapshots of a peer-to-peer network are predicted and used to determine that a computing device may be better off waiting until a future point in time to download specific digital content. For another example, computing devices are mapped into different groups based on location information, and inter-group information is used to identify other computing devices for a computing device to send download requests for digital content. For a further example, information indicative of scarcity associated with different digital content units is used to prioritize distribution of the digital content units.

Abstract: Embodiments are disclosed for presenting a digital content item comprising a plurality of content portions. One example embodiment includes a computing device comprising a primary content storage machine, where the primary content storage machine is configured to selectively store one or more content portions of a digital content item. The computing device is configured to determine a dynamically changing content access window including one or more content portions useable to provide an above-threshold user experience based on a current access position of the digital content item. The computing device is configured to dynamically load the primary content storage machine with the content portions of the digital content item corresponding to the content access window and dynamically unload the content portions of the digital content item outside of the content access window from the primary content storage machine.

Abstract: Embodiments are disclosed that relate to generating digital content recommendations for a user based upon how the user accesses the assets of a digital content item. For example, one disclosed embodiment provides a method including receiving from a remote computing device content access information regarding an order in which content portions of a selected digital content item were accessed by the user, and, storing the content access information. The method further includes comparing the content access information for the user to content access information for other users that consumed the selected digital content item to determine other users with similar content access patterns, and sending digital content recommendations to the user based upon content consumption information for the other users.

Abstract: Various embodiments are disclosed that relate to buffering digital content items in different orders for different user experiences. For example, one disclosed embodiment provides, on a computing device, a method for providing a buffering order for a digital content item. The method includes receiving from a remote computing device a request to access a selected digital content item, the selected digital content item comprising a plurality of content portions consumable in a plurality of different orders, the plurality of different orders corresponding to a plurality of user experiences for the selected digital content item, and in response, providing a selected content provision schema selected from a plurality of content provision schemas for the selected digital content item, each content provision schema defining a buffering order of the plurality of content portions of the selected digital content item for a corresponding user experience of the selected digital content item.

Abstract: Embodiments are disclosed that relate to buffering portions of a digital content item in different orders for different users. For example, one disclosed embodiment provides a method of providing a buffer ordering for a digital content item. The method includes tracking content access information for each user of a plurality of users, the content access information for each user comprising information regarding how content portions of each of one or more digital content items were accessed by the user. The method further comprises providing a different buffer ordering for a first user of a selected digital content item than for a second user based upon the content access information.

Abstract: Systems and methods for personalized media recommendation are provided. One example system may include a media server including a media database containing a plurality of media tracks and a database schema defining multiple media styles. The media server may further include a media delivery engine configured to transmit media tracks to each of a plurality of media clients. The system may further include a feedback module configured to receive history data from each of the media clients. Further still, the system may include a recommendation engine configured to compute a plurality of user-specific style weights, and to select one or more recommended media tracks for each user, The recommendation engine may be further configured to send a recommendation message to the media client including media links to the recommended media tracks on the media server, to cause the media client to display the media links to a user.

Abstract: Personalized media charts are described. In embodiments, a request can be received for a personalized media chart that lists media assets as recommendations in a relevant order for a user. The media assets are initially ordered in the personalized media chart according to a chart rank for each media asset. A relevance factor is then determined for each of the media assets, where the relevance factor for a media asset indicates a relevance of the media asset to the user. A chart rank adjustment is calculated for each of the media assets based on the relevance factors for each of the media assets, and the media assets are then reordered in the personalized media chart according to the chart rank adjustments for each of the media assets.

Abstract: Systems and methods for personalized media recommendation are provided. One example system may include a media server including a media database containing a plurality of media tracks and a database schema defining multiple media styles. The media server may further include a media delivery engine configured to transmit media tracks to each of a plurality of media clients. The system may further include a feedback module configured to receive history data from each of the media clients. Further still, the system may include a recommendation engine configured to compute a plurality of user-specific style weights, and to select one or more recommended media tracks for each user, The recommendation engine may be further configured to send a recommendation message to the media client including media links to the recommended media tracks on the media server, to cause the media client to display the media links to a user.