Surrogate Language Model Surveying for Faster Response Collection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing questionnaire survey systems require time and cost to acquire respondents and often face reluctance from participants to answer surveys.
Innovation Solution
An information processing apparatus and method utilizing a vicariously answering language model trained by machine learning to generate answers to questions on behalf of users, facilitating survey collection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional questionnaire survey systems are used to collect survey responses, then survey data can be gathered, but it requires significant time and cost to acquire respondents and participants are reluctant to spend time answering
Solution Approach 1:
The patent uses a language model to generate synthetic responses that copy the style and content characteristics of real user responses. The model is trained on actual survey data and can generate answers that statistically resemble human responses, allowing the system to collect survey data without requiring actual human participants to spend time answering questions.
Solution Approach 2:
The language model performs self-service by automatically generating survey responses without human intervention. The system trains the model on historical survey data and then uses the trained model to autonomously generate responses to new survey questions, eliminating the need for human respondents to manually answer questions and significantly reducing data collection time.
2Productivity
If traditional questionnaire survey systems are used, then survey data can be collected, but acquiring respondents requires significant time and cost
Solution Approach 1:
The language model creates synthetic survey responses that copy the statistical properties and content patterns of real human responses. By generating responses that statistically resemble actual user answers, the system can obtain the necessary survey data quantity without requiring a corresponding quantity of human respondents, thereby reducing both time and cost.
3Reliability
If respondents are asked to answer questionnaires manually, then authentic user feedback is obtained, but respondents are unwilling to spend time on surveys
Solution Approach 1:
The language model incorporates feedback loops where it continuously refines its response generation based on the quality and patterns of survey responses. The model learns from actual user feedback data and adjusts its generation parameters to better match authentic user responses, maintaining reliability while eliminating the need for human participants.
Solution Approach 2:
The system copies the essential characteristics of authentic user responses through the language model rather than requiring actual user participation. The model is trained on real survey data and generates responses that replicate the statistical patterns, tone, and content characteristics of genuine user feedback, maintaining response authenticity without requiring respondents to actively participate.
Data Source
AI summary
Easily conducting a survey in which answers to a predetermined question are collected is made possible. An information processing apparatus includes: an accepting section for accepting a request for a survey in which answers to a predetermined question are collected; and a negotiating section for conducting, with a predetermined negotiating partner, negotiations for causing a vicariously answering language model to generate an answer to the predetermined question, the vicariously answering language model having been trained by machine learning so as to be capable of generating, as a surrogate for a predetermined user, an answer to an inputted question. This information processing apparatus makes it possible to optimize respondents of a question.