Crowdsourced Action Recipes for Digital Assistant Command Execution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional digital assistants face issues such as privacy concerns, frequent misinterpretations of spoken commands, unavailability due to weak signals, and the need for specific syntax, limiting their functionality across multiple applications and within mobile devices.
Innovation Solution
A crowdsourced digital assistant system that allows users to record and execute 'action recipes' – sequences of actions across multiple applications using customizable commands, enabling efficient execution of actions through a dataset stored on user devices and an action cloud server.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If conventional digital assistants use speech recognition and natural language processing, then they can interpret commands, but they frequently misinterpret spoken commands and require specific syntax
Solution Approach 1:
The patent segments the command execution into two distinct phases: (1) intent recognition phase where the digital assistant identifies the user's goal using natural language processing, and (2) action recipe retrieval phase where pre-defined action sequences are matched and executed. This segmentation allows the system to separate interpretation from execution, improving reliability by using pre-validated action recipes rather than executing interpreted commands directly.
Solution Approach 2:
The system performs preliminary action by pre-defining and storing action recipes that map intended actions to specific sequences of operations across applications. These action recipes are created in advance and stored in a database, allowing the digital assistant to retrieve and execute proven action sequences rather than interpreting and executing commands in real-time, thereby reducing misinterpretation errors.
2Adaptability or versatility
If digital assistants execute actions within applications, then functionality is improved, but they cannot execute multiple functions across different applications simultaneously
Solution Approach 1:
The patent implements universality by creating a universal action recipe framework that can execute multiple different actions across multiple different applications through a single integrated system. The action recipe database stores diverse action sequences (e.g., sending messages, making calls, controlling smart home devices) that can be retrieved and executed uniformly, allowing one digital assistant system to perform many functions without requiring separate integration mechanisms for each application.
Solution Approach 2:
The system introduces an intermediary layer in the form of action recipes that mediate between the user's high-level intentions and the low-level application-specific operations. Action recipes serve as translators that convert user commands into sequences of application-specific actions, eliminating the need for direct complex integrations between the digital assistant and each individual application while still enabling multi-application functionality.
3Measurement precision
If users need to structure commands in specific dialects, then command precision is improved, but user comfort and ease of use deteriorate
Solution Approach 1:
The patent inverts the traditional command processing approach by instead of requiring users to structure commands precisely, the system provides precise action sequences and asks users to provide natural, comfortable commands. The action recipes contain the structured, application-specific action sequences, while the user interface accepts loose, natural language inputs that are then mapped to the appropriate action recipes, effectively inverting the burden from the user to the system.
Data Source
AI summary
Embodiments of the present invention are directed to action recipes for a crowdsourced digital assistant. Users can define an action recipe by recording a set of inputs across one or more applications, by providing multiple sub-commands in a single on-the-fly command, by providing one or more associated commands, or otherwise. An action recipe dataset is generated, and stored and indexed on a user device and/or on an action cloud server. As such, any user can invoke an action recipe by providing an associated command to a crowdsourced digital assistant application on a user device. The crowdsourced digital assistant searches for a matching command on the user device and/or the action cloud server, and if a match is located, the corresponding action recipe dataset is accessed, and the crowdsourced digital assistant emulates the actions in the action recipe on the user device.