Abstract
This document describes a technique for processing voice commands received by a voice activated computing system to identify search requests and trigger keywords in the voice commands. Based on the search requests and the trigger keywords, the system can identify search entities, and associated actions. The search entity and the actions can be determined, for example, based on techniques such as heuristics and knowledge graphs. The system can then form one or more search entity-action pairs. Third-party content providers can compete to provide their content to the user based not on the search request alone, but on the search entity-action pair. That is, the third-party content providers can enter an auction and bid for providing their content over others. The selected third-party content can then be provided to the user. In some instances, each content available for a search entity-action pair can have an associated quality score, which can be representative of a likelihood of the user accepting the content if presented to the user. Based on a combination of the auction bid and the quality score, the system can select a content item and generate an audio file that provides the content item to the user.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Lanza, John D., "Delivery Of Content Based on Search Entity-Action Pairs In a Voice Activated Computing System", Technical Disclosure Commons, (July 03, 2017)
https://www.tdcommons.org/dpubs_series/574