Abstract: Technologies are described herein for providing dynamic natural language interactions between a user and a computing device. In one aspect, a computing device managing a conversation with a user is enhanced with the identification and management of one or more topics. Using techniques described herein, the computing device can focus on one or more topics, shift between topics and/or introduce new topics. Techniques disclosed herein may also manage and process interruptions that may be introduced during a conversation. Dynamic access of contextual information may assist in the generation of contextually-relevant statements, and the contextual information may be used to balance priorities between various topics. Each topic may also have an associated decay rate so that the lifespan of individual topics may track realistic scenarios. In addition, the priorities of individual topics may be dynamically adjusted so topics may track events created by a user and a computing device.

Abstract: Technologies are described herein for providing dynamic natural language interactions between a user and a computing device. In one aspect, a computing device managing a conversation with a user is enhanced with the identification and management of one or more topics. Using techniques described herein, the computing device can focus on one or more topics, shift between topics and/or introduce new topics. Techniques disclosed herein may also manage and process interruptions that may be introduced during a conversation. Dynamic access of contextual information may assist in the generation of contextually-relevant statements, and the contextual information may be used to balance priorities between various topics. Each topic may also have an associated decay rate so that the lifespan of individual topics may track realistic scenarios. In addition, the priorities of individual topics may be dynamically adjusted so topics may track events created by a user and a computing device.