Abstract: Systems and methods related to engaging with a virtual assistant via ancillary input are provided. Ancillary input may refer to non-verbal, non-tactile input based on eye-gaze data and/or eye-gaze attributes, including but not limited to, facial recognition data, motion or gesture detection, eye-contact data, head-pose or head-position data, and the like. Thus, to initiate and/or maintain interaction with a virtual assistant, a user need not articulate an attention word or words. Rather the user may initiate and/or maintain interaction with a virtual assistant more naturally and may even include the virtual assistant in a human conversation with multiple speakers. The virtual assistant engagement system may utilize at least one machine-learning algorithm to more accurately determine whether a user desires to engage with and/or maintain interaction with a virtual assistant. Various hardware configurations associated with a virtual assistant device may allow for both near-field and/or far-field engagement.

Abstract: Automatic extraction and application of conditional tasks from content is provided. A conditional task system includes a classifier that is trained and used to identify conditional tasks and to learn appropriate times and methods to engage a user for reminding the user about conditional tasks. The conditional task system includes components for enabling an automated detection of a conditional task, extracting of attributes that characterize a condition associated with the task, using information about the condition to determine how to monitor for satisfaction of the condition, determining when and how to engage the user about the task, and notifying the user at an appropriate time and using an appropriate method when the condition is satisfied.

Abstract: Recommending one or more relevant skills to a user that are appropriate for the user's current context is provided. A machine learning contextual skills model is trained on data associated with skills that are invoked by the particular user and by various users and context data that defines characteristics of the context of the users, the users' computing devices, and the users' environments in association with the invoked skills. Responsive to a trigger event associated with the user and a computing device, current context data and user data is collected and passed to the contextual skills system. The contextual skills model is applied to the collected data and generates a ranked list of skills. One or more top-ranking skills are selected and provided to the user as one or more skill recommendations that are determined to be relevant to the user and relevant to the user's current context.

Abstract: Generating talking points on behalf of a user and providing the talking points to the user is provided. Talking points can help users to quickly and easily start a personalized conversation that enables participants of the conversation to make a connection, thus eliciting better communication and collaboration. Talking points can be generated based on commitments or requests identified in a previous conversation, anomalies or patterns identified in context data, or information items of interest collected from various data sources. One or more top-ranking talking points are surfaced to the user to help the user start a new conversation or to continue a previous conversation.

Abstract: Technologies pertaining to automatically identifying sets of query attribute values that are highly correlative with user dissatisfaction with a search engine are described. Dissatisfied queries are automatically identified through analysis of search logs, wherein a dissatisfied query is a query submitted to a search engine by a user, wherein the user was dissatisfied with search results provided by the search engine responsive to receipt of the query. Sets of query attribute values that are highly correlated with dissatisfied queries, and thus user dissatisfaction, are automatically identified based at least in part upon the identifying of the dissatisfied queries. Subsequent to identifying a set of query attribute values, a segment-specific ranker is learned that is configured to rank search results responsive to receipt of a query with the set of query attribute values, wherein the segment-specific ranker outperforms a general purpose ranker for queries having the set of query attribute values.

Abstract: Biases in search and retrieval (i.e., situations where searchers seek or are presented with information that significantly deviates from the truth) may be detected by comparison to one or more authoritative sources. Once bias or potential bias is detected, techniques may be applied to indicate and/or compensate for the bias. Such techniques may allow users to more easily assess the veracity of search results, and increase the chances that users will locate accurate answers to their questions.

Abstract: Biases in search and retrieval (i.e., situations where searchers seek or are presented with information that significantly deviates from the truth) may be detected by comparison to one or more authoritative sources. Once bias or potential bias is detected, techniques may be applied to indicate and/or compensate for the bias. Such techniques may allow users to more easily assess the veracity of search results, and increase the chances that users will locate accurate answers to their questions.

Abstract: Biases in search and retrieval (i.e., situations where searchers seek or are presented with information that significantly deviates from the truth) may be detected by comparison to one or more authoritative sources. Once bias or potential bias is detected, techniques may be applied to indicate and/or compensate for the bias. Such techniques may allow users to more easily assess the veracity of search results, and increase the chances that users will locate accurate answers to their questions.

Abstract: Examples are provided for a health monitoring computing system, including a data interface configured to receive appointment data and health tracking data, the health tracking data including biometric data measured by one or more biometric sensors worn by a user, health-scheduling logic configured to determine that the user is to see a healthcare professional within a threshold period of time based at least on computer analysis of the appointment data, and appointment-optimization logic configured to generate a list of diagnosis-relevant health information based at least on computer analysis of the health tracking data. The example health monitoring computing system also includes an output device configured to present the list to the user for editing prior to an appointment with the healthcare professional.

Abstract: Examples are disclosed that relate to providing healthcare-related information. One example provides a computing device comprising a logic machine and a storage machine holding instructions executable by the logic machine to receive an input of information regarding a health state of a user, obtain, based upon the information regarding the health state of the user, an inference of a possible health condition of the user, output a notification of the inference, the notification comprising a first representation of the inference, receive data representing a mechanism for authorizing a healthcare practitioner to access a second representation of the inference, and output a user-selectable control for triggering the mechanism. The instructions may be further executable to receive an input via the user-selectable control triggering the mechanism, and, in response, send authorization to provide the healthcare practitioner with access to the second representation of the inference.

Abstract: Determining stress and productivity insights based on computerized data is described. Productivity data and stress-related behavior data associated with a user may be determined. The stress-related behavior data may be based on sensor data received from sensor(s) associated with a device corresponding to the user. The stress-related behavior data may be utilized to determine a first value indicative of stress associated with the user and the productivity data may be utilized to determine a second value indicative of productivity of the user. The productivity data, the stress-related behavior data, the first value, and/or the second value may be utilized to determine a recommendation. The recommendation may be intended to modify the first value and/or the second value. A user interface configured to communicate the recommendation to the user may be presented via the device.

Abstract: A wearable device includes one or more biometric sensors configured to determine biometric data of a wearer of the wearable device and a stress assessment tool. The stress assessment tool is configured to determine a periodic stress score, via a machine-learning model, based at least on the biometric data, visually present, via a display associated with the wearable device, a graphical user interface (GUI) including the periodic stress score, receive wearer feedback evaluating the accuracy of the periodic stress score, adjust the machine-learning model based on the wearer feedback, determine a reassessed periodic stress score, via the machine-learning model, and visually present, via the display, the reassessed periodic stress score in the GUI.

Abstract: Examples are disclosed herein that relate to correlating health data outcomes to work data of a user and recommending future work activities based upon the correlation. One example provides a computing device configured to monitor health data relating to health behavior of a user over time, determine a baseline value for a health outcome of the user based on the monitored health data, and receive work data relating to a work schedule of the user. Responsive to the health outcome deviating from the baseline value, the computing device is configured to determine that the deviation in the health outcome correlates with the work data, and based at least upon determining that the deviation in the health outcome correlates with the work data, output a notification of the determination to the user.

Abstract: A method for generating productivity insights includes receiving health data for a user of a productivity evaluation service. From the health data, health behaviors and health effects of the user are determined. Productivity data for the user is received, and from the productivity data, productivity behaviors and productivity effects of the user are determined. Associations between changes in the health data and changes in the productivity data are identified. Based on one of the associations, a productivity insight is generated for the user including a prompt to engage in a health behavior that is associated with a desirable productivity effect.

Abstract: Examples are disclosed herein that relate to modifying an analysis of personal behavior based on determining a subset of personal data to be non-routine. One example provides a computing device configured to receive personal data relating to personal behavior of a user, receive contextual data regarding the personal data, determine a subset of the personal data to be non-routine based upon the contextual data, and modify an analysis of personal behavior based upon the subset of the personal data determined to be non-routine.

Abstract: Examples are disclosed herein that relate to integrating health data and calendar data of one or more users and providing insights for a selected user to help the user accomplish an outcome of interest. The insights may be identified based on a group of cohorts determined to be similar to the selected user and/or used to predict a likelihood that the selected user will achieve an outcome of interest. Additional insights may be provided by monitoring an effect that following a recommendation has on the selected user achieving the outcome of interest. Recommendations and/or updates to recommendations may be provided based on the insights.

Abstract: Examples are disclosed herein that relate to integrating health data and work data of one or more users and recommending a time to perform a health activity. One example provides a computing device configured to receive health data relating to health behavior, the health data comprising information regarding a relationship between health activity scheduling and health outcome, receive work data relating to work activities, compare the health data and the work data to determine a time at which to recommend performing a health activity, and output a recommendation regarding the time to perform the health activity.

Abstract: A device includes a memory that stores a prefetching model. A control module receives a content page including one or more links each associated with selectable content and collects data associated with the content page. The collected data includes at least one of first data indicative of respective relationships between each of the links and a viewport of the device and second data indicative of characteristics of the viewport. The control module further assigns, using the prefetching model, respective scores to each of the links based on the collected data, and selectively generates, based on the assigned scores, a request to prefetch the selectable content associated with at least one of the links.

Abstract: A computing system comprises a client computing device configured to execute a personal assistant application program. The personal assistant application program is configured to receive user data from interaction of a user with the client computing device, user interaction with additional devices, or system networked to the client computing device, to sense a user condition based on the user data received, to analyze the user condition to identify a user health issue, present, via a user interface associated with the client computing device, a suggestion for the user to treat, overcome or improve the user health issue, assess a degree to which the user has followed the suggestion, and modify subsequent suggestions to the user based on the degree to which the suggestion was followed.

Abstract: Aspects of the technology described herein provide a more efficient user interface by providing suggestions that are tailored to a specific user's interests. The suggestions may be provided by a personal assistant or some other application running on a user's computing device. A goal of the technology described herein is to provide relevant suggestions when the user can and actually wants to use them. The suggestions are designed to provide information or services the user wants to use.