Abstract: Aspects of the present disclosure provide accelerated transfer learning for machine learning models, such as neural networks, as a service. In doing so, the aspects disclosed herein decouple the conversion process for use of machine learning models from deployment of the models in a production environment through the creation of one or more common generic models that execute on a hardware accelerator, such as a FPGA. Additional aspects of the disclosure relate to the creation of an accelerated machine learning model. The accelerated machine learning model is created by identifying common portions of a machine learning model that can be leveraged by a plurality of different machine learning models. Still further aspects of the disclosure relate to training scenario specific machine learning models for use with an accelerated machine learning model.

Abstract: Computerized systems are provided for automatically determining action items of an event, such as a meeting. The determined action items may be personalized to a particular user, such as a meeting attendee, and may include contextual information enabling the user to understand the action item. In particular, a personalized action item may be determined based in part from determining and utilizing particular factors in combination with an event dialog, such as an event speaker's language style; user role in an organization; historical patterns in communication; event purpose, name, or location; event participants, or other contextual information. Particular statements are evaluated to determine whether the statement likely is or is not an action item. Contextual information may be determined for action items, which then may be provided to the particular user during or following the event.

Abstract: Personalized notifications are provided to a user. The notifications correspond to events, which can include information items, unaddressed or unnoticed by a user. The unaddressed events may be determined based on sensor data provided by a user device associated with the user including user activity pattern information. Further, an urgency level of the unaddressed event and user availability for responding to the event may be determined. The availably may comprise a model with score(s) corresponding to the user's capability for carrying out tasks associated with addressing the event. Additionally, notification content may be generated based on information about availability and the unaddressed event, and used to provide notification(s). The notification content may include a recommendation to present a notification to the user at a time determined likely to be convenient for the user to be notified of the event or where the user likely has capacity to address the event appropriately.

Abstract: Personalized notifications are provided to a user. The notifications correspond to events, which can include information items, unaddressed or unnoticed by a user. The unaddressed events may be determined based on sensor data provided by a user device associated with the user including user activity pattern information. Further, an urgency level of the unaddressed event and user availability for responding to the event may be determined. The availably may comprise a model with score(s) corresponding to the user's capability for carrying out tasks associated with addressing the event. Additionally, notification content may be generated based on information about availability and the unaddressed event, and used to provide notification(s). The notification content may include a recommendation to present a notification to the user at a time determined likely to be convenient for the user to be notified of the event or where the user likely has capacity to address the event appropriately.

Abstract: Personalized reminder notifications are provided regarding communication-related events. The reminders correspond to missed communications or events unaddressed by a user, and may be determined based on sensor data provided by a user device. Further, an urgency level for the unaddressed event and a user availability model for responding to the event may be determined. The model may include a set of scores corresponding to the user's capability for carrying out tasks associated with responding to the unaddressed event, such as composing email or initiating a call. Additionally, reminder notification content may be generated based on information about availability and the unaddressed event, and used to provide notification(s). The notification content may include a recommendation to present a notification to the user at a future time determined likely to be convenient for the user to respond to the event or where the user likely will have capacity to respond appropriately.

Abstract: Personalized notifications are provided to a user. The notifications correspond to events, which can include information items, unaddressed or unnoticed by a user. The unaddressed events may be determined based on sensor data provided by a user device associated with the user including user activity pattern information. Further, an urgency level of the unaddressed event and user availability for responding to the event may be determined. The availably may comprise a model with score(s) corresponding to the user's capability for carrying out tasks associated with addressing the event. Additionally, notification content may be generated based on information about availability and the unaddressed event, and used to provide notification(s). The notification content may include a recommendation to present a notification to the user at a time determined likely to be convenient for the user to be notified of the event or where the user likely has capacity to address the event appropriately.

Abstract: A logical location of a user may be inferred using semantics of the user's computing device(s). The logical locations may correspond to venues visited by the user, such as frequently visited venues, referred to as hubs. Logical hubs, representing logical locations, may be tagged to the computing device by classifying the device as being associated with a logical hub. Classification may be based on signals or features of the device, such as device usage or device characteristics. The device may be monitored to detect user activity. Based on analysis of the user activity and associated logical hub, the user's logical location may be inferred. A computer user-experience may be personalized to the user based on the user's inferred logical location or logical hub(s) associated with the device. In some cases, geographical location information is not used to determine the logical hubs or infer a user's logical location.

Abstract: Personalized reminder notifications are provided regarding communication-related events. The reminders correspond to missed communications or events unaddressed by a user, and may be determined based on sensor data provided by a user device. Further, an urgency level for the unaddressed event and a user availability model for responding to the event may be determined. The model may include a set of scores corresponding to the user's capability for carrying out tasks associated with responding to the unaddressed event, such as composing email or initiating a call. Additionally, reminder notification content may be generated based on information about availability and the unaddressed event, and used to provide notification(s). The notification content may include a recommendation to present a notification to the user at a future time determined likely to be convenient for the user to respond to the event or where the user likely will have capacity to respond appropriately.

Abstract: Systems, methods, and computer-readable storage media are provided for tailoring or customizing computing experiences based upon contextual signals. One or more contextual signals associated with a user of a computing device are detected. Based upon the detected contextual signal(s), a probable action of the user is predicted, for instance, taking into account historical actions of the user and/or cumulative historical actions of a plurality of users as such actions relate to the detected contextual signal(s). At least one user interface, such as an action menu is automatically (that is, without affirmative user interaction) customized to include a menu item that can activate a function to help complete the probable action.

Abstract: A logical location of a user may be inferred using semantics of the user's computing device(s). The logical locations may correspond to venues visited by the user, such as frequently visited venues, referred to as hubs. Logical hubs, representing logical locations, may be tagged to the computing device by classifying the device as being associated with a logical hub. Classification may be based on signals or features of the device, such as device usage or device characteristics. The device may be monitored to detect user activity. Based on analysis of the user activity and associated logical hub, the user's logical location may be inferred. A computer user-experience may be personalized to the user based on the user's inferred logical location or logical hub(s) associated with the device. In some cases, geographical location information is not used to determine the logical hubs or infer a user's logical location.

Abstract: Personalized reminder notifications are provided regarding communication-related events. The reminders correspond to missed communications or events that are unaddressed by a user, and may be determined based on sensor data provided by a user device. Further, an urgency level for the unaddressed event and a user availability model for responding to the event also may be determined. The model may include a set of scores corresponding to the user's capability for carrying out tasks associated with responding to the unaddressed event, such as composing email or initiating a call. Additionally, reminder notification content may be generated based on information about availability and the unaddressed event, and used to provide notification(s). The notification content may include a recommendation to present a notification to the user at a future time determined likely to be convenient for the user to respond to the event or where the user likely will have capacity to respond appropriately.

Abstract: Personalized reminder notifications are provided regarding communication-related events. The reminders correspond to missed communications or events that are unaddressed by a user, and may be determined based on sensor data provided by a user device. Further, an urgency level for the unaddressed event and a user availability model for responding to the event also may be determined. The model may include a set of scores corresponding to the user's capability for carrying out tasks associated with responding to the unaddressed event, such as composing email or initiating a call. Additionally, reminder notification content may be generated based on information about availability and the unaddressed event, and used to provide notification(s). The notification content may include a recommendation to present a notification to the user at a future time determined likely to be convenient for the user to respond to the event or where the user likely will have capacity to respond appropriately.

Abstract: Personalized notifications are provided to a user. The notifications correspond to events, which can include information items, unaddressed or unnoticed by a user. The unaddressed events may be determined based on sensor data provided by a user device associated with the user including user activity pattern information. Further, an urgency level of the unaddressed event and user availability for responding to the event may be determined The availably may comprise a model with score(s) corresponding to the user's capability for carrying out tasks associated with addressing the event. Additionally, notification content may be generated based on information about availability and the unaddressed event, and used to provide notification(s). The notification content may include a recommendation to present a notification to the user at a time determined likely to be convenient for the user to be notified of the event or where the user likely has capacity to address the event appropriately.