Abstract: A computing device includes a logic machine and a storage machine holding instructions. The instructions are executable by the logic machine to, receive a machine-understandable representation of voice input, determine a pronoun associated with the voice input, map the pronoun to an object of user focus, generate one or more focus-specific search queries for the object of user focus, execute the one or more focus-specific search queries, and present the results of the one or more focus-specific search queries.

Abstract: A computing device includes a logic machine and a storage machine holding instructions. The instructions are executable by the logic machine to, receive a machine-understandable representation of voice input, determine a pronoun associated with the voice input, map the pronoun to an object of user focus, generate one or more focus-specific search queries for the object of user focus, execute the one or more focus-specific search queries, and present the results of the one or more focus-specific search queries.

Abstract: Representative embodiments disclose mechanisms to extract an emotional state from contextual user data and public use data collected from one or more devices and/or services. The contextual and public data are combined into an enriched data set. An emotional model, tailored to the user, extracts an emotional state form the enriched data set based on one or more machine learning techniques. The emotional state is used to identify one or more actions that change operation of one or more devices and/or services in order to achieve a change in emotional state, compatibility between the emotional state and device and/or service interaction with the user, or both. Implicit and/or explicit feedback is collected and used to change prediction of the emotional state and/or selection of the actions.