Survey Question Domain Matching With ML Model Arbitration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing survey systems face challenges in generating appropriate responses due to manual effort, user errors, and difficulty in associating questions with corresponding answers, especially when different organizations use differently phrased or formatted surveys, leading to variance and omission of relevant information.
Innovation Solution
A data processing system employs machine learning models to predict question domains, arbitrate predictions, and match questions with corresponding answers, allowing for user feedback to improve model training and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual question responses are generated, then users can provide customized answers, but substantial time and duplicative effort are incurred
Solution Approach 1:
The system enables automated self-service for survey response generation by using machine learning models to automatically match questions with appropriate answers from a knowledge base, eliminating the need for manual response creation while maintaining customization and accuracy
Solution Approach 2:
The patent replaces the manual mechanical process of writing and matching survey responses with an automated computational system that uses natural language processing and machine learning algorithms to perform question-answer matching, significantly improving efficiency
2Productivity
If banked responses are used, then response generation time is reduced, but manual searching and user error in selection occur
Solution Approach 1:
The system replaces manual searching and selection of banked responses with an automated machine learning-based matching system that computes similarity between questions and potential answers, eliminating user error while maintaining fast response retrieval
Solution Approach 2:
The patent introduces machine learning models as an intermediary between the question and the banked responses, automatically computing matches and presenting ranked results to users, thereby reducing both manual effort and selection errors
3Measurement precision
If multiple machine learning models are used for domain prediction, then prediction accuracy is improved, but system complexity increases
Solution Approach 1:
The patent combines multiple machine learning models into a unified ensemble system that aggregates their predictions through voting or averaging, improving domain prediction accuracy while managing complexity through integrated architecture and standardized interfaces
Data Source
AI summary
The present solution provides systems and methods to receive a data set comprising a representation of one or more questions from a survey and provide the data set as input to each of a plurality of machine learning models trained to predict a domain associated with the one or more questions. The systems and methods can receiving as output a first domain prediction for the domain from each of the plurality of machine learning models and determine a second domain prediction for the domain for each question of the one or more questions based on applying a function to each of the first domain predictions. The systems and methods can select, based on the data set and the second domain prediction, an enumerated list of one or more answers from an answer set and cause a display of the enumerated list via a user interface for a selection.