Natural Language Task Automation via Machine Learning Step Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users face inefficiencies when performing online tasks as they need to manually execute multiple steps, such as logging into accounts or making purchases, which can be time-consuming and cumbersome.
Innovation Solution
A system utilizing machine learning systems to generate and perform specific steps based on general natural language commands, allowing users to input tasks like 'purchase items in my cart' and automatically executing the necessary actions on a webpage, such as selecting the cart icon, entering credit card information, and confirming the purchase.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If users manually perform multiple steps to complete online tasks (such as logging in, entering information, selecting options), then the system can accurately execute each step, but the user spends significant time and effort
Solution Approach 1:
The system performs tasks automatically without requiring user intervention for each step. The electronic processor executes multiple steps autonomously based on a single user input, allowing the system to serve itself by completing routine operations like form filling, navigation, and data entry without human assistance
Solution Approach 2:
The patent breaks down a general natural language command into multiple specific executable steps. The system segments the overall task into discrete actions (e.g., navigate to URL, fill form field, click button) that can be individually processed and executed, transforming one complex task into manageable step-by-step operations
2Ease of operation
If the system automates task execution by generating multiple specific steps from a general command, then user effort is reduced, but the system complexity increases due to multiple machine learning systems
Solution Approach 1:
The patent introduces a natural language processing intermediary layer between the user and the task execution system. This intermediary translates user-friendly natural language commands into structured step-by-step instructions, shielding users from the underlying system complexity while enabling automated execution through multiple machine learning systems
3Extent of automation
If the system uses multiple machine learning systems to determine steps and execute tasks, then task automation capability improves, but the training and maintenance burden increases
Solution Approach 1:
The system performs preliminary training of multiple machine learning systems using gathered training data before actual task execution. By pre-training the models with example tasks and outcomes, the system prepares the machine learning systems in advance to handle various tasks autonomously, reducing the maintenance burden during operational phases
Data Source
AI summary
A system for performing a task with multiple specific steps given a general natural language command. The system includes an electronic processor. The electronic processor is configured to receive a general natural language command specifying a task to perform and, using a first machine learning system, generate a plurality of specific steps associated with the general natural language command. The electronic processor is also configured to, using the plurality of specific steps and a second machine learning system, perform the task, determine whether the task is performed successfully, and, when the task is not performed successfully, retrain the first machine learning system, second machine learning system, or both.