Contextual attention across diverse artificial intelligence voice assistance systems

A system for using artificial intelligence voice assistants comprising:

• a hardware processor;
• and
• a memory that stores a computer program product, the computer program product when executed by the hardware processor, causes the hardware processor to:
• register a user to a common distributed computing platform that provides access to a plurality of diverse voice assistants, wherein each diverse voice assistant is assigned to a vendor location and are provided by different service providers;
• receive at least one first command issued by a user to a first voice assistant of the plurality of diverse voice assistants at a first location;
• build a context database by storing the at least one first command on the common distributed computing platform;
• receive at least one second command issued by the user to a second voice assistant of the plurality of diverse voice assistants at a second location;
• analyze the at least one second command in comparison with the at least one first command on the context database, with a context assigning engine that utilizes a pre-attention bi-directional long short term memory (LSTM) model, an attention layer, and a post-attention bi-directional LSTM model of the common distributed computing platform, wherein similarity between the at least one first command and the at least one second command provides context;
• send the context from the context assigning engine receiving to the second voice assistant;
• and
• employ the context to answer the at least one second command.

A computer program product for using artificial intelligence voice assistants comprising a computer readable storage medium having computer readable program code embodied therewith, the computer readable program code executable by a processor to cause the processor to:

• register, using the processor, a user to a common distributed computing platform that provides access to a plurality of diverse voice assistants, wherein each voice assistant is assigned to a vendor location and are provided by different service providers;
• receive, using the processor, at least one first command issued by a user to a first voice assistant of the plurality of diverse voice assistants at a first location;
• build, using the processor, a context database by storing the at least one first command on the common distributed computing platform;
• receive, using the processor, at least one second command issued by the user to a second voice assistant of the plurality of diverse voice assistants at a second location;
• analyze, using the processor, the at least one second command in comparison with the at least one first command on the context database, with a context assigning engine that utilizes a pre-attention bi-directional long short term memory (LSTM) model, an attention layer, and a post-attention bi-directional LSTM model of the common distributed computing platform, wherein similarity between the at least one first command and the at least one second command provides context;
• send, using the processor, the context from the context assigning engine receiving to the second voice assistant;
• and
• employ, using the processor, the context to answer the at least one second command.