Agent Utterance Probability Level Update System
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for supervising agent utterance data lack flexibility in handling the character of the agent, leading to inefficient processing and the inability to achieve a rich set of utterance data, as they only allow binary decisions on removing prohibited language without considering the agent's personality.
Innovation Solution
An information processing system that updates an utterance probability level based on user feedback, allowing for flexible auditing and recording of the likelihood of the agent uttering specific content, enabling more efficient and personalized management of agent utterance data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If binary decision system is used for supervising agent utterance data, then prohibited language can be removed, but flexibility in handling agent character is lost
Solution Approach 1:
The patent changes the parameter from binary (remove/keep) to continuous probability values (0-1), allowing nuanced control over utterance selection. This enables the system to express degrees of appropriateness rather than simple yes/no decisions, resolving the contradiction between reliability and adaptability.
Solution Approach 2:
The system dynamically adjusts utterance probability levels based on feedback from multiple users and supervisors. The probability values are not fixed but evolve over time through continuous updates, allowing the system to adapt to changing requirements while maintaining reliable filtering.
2Manufacturing precision
If specialist supervision is used for utterance data, then quality control is improved, but work efficiency decreases due to aggregated processing units
Solution Approach 1:
The patent merges multiple supervision sources (general users and specialists) into a unified feedback system. Their inputs are combined to update utterance probability levels, allowing specialists to focus on quality control while distributing initial filtering to general users, thus improving both quality and efficiency.
Solution Approach 2:
The system enables self-service through automated probability level updates based on collected feedback. Rather than requiring specialists to manually review every utterance, the system automatically adjusts probabilities based on user interactions and supervisor inputs, significantly improving productivity while maintaining quality.
3Ease of manufacture
If aggregated data processing is used, then work scheduling is simplified, but rich set of utterance data cannot be achieved
Solution Approach 1:
The system operates continuously by collecting feedback from users in real-time and automatically updating utterance probability levels. This continuous process allows the database to grow rich and diverse without requiring periodic batch processing, maintaining both ease of scheduling and data richness.
Solution Approach 2:
The system performs preliminary filtering and probability assessment automatically before specialist review. This preliminary action prepares data in advance, allowing specialists to focus only on borderline cases, thus simplifying scheduling while enabling comprehensive data collection.
Data Source
AI summary
The present disclosure provides an information processing system and an information processing method capable of auditing the utterance data of an agent more flexibly. In one example, an information processing system includes: a storage section that stores utterance data of an agent; a communication section that receives request information transmitted from a client terminal and requesting utterance data of a specific agent from a user; and a control section that, when the request information is received through the communication section, replies to the client terminal with corresponding utterance data, and in accordance with feedback from the user with respect to the utterance data, updates an utterance probability level expressing a probability that the specific agent will utter utterance content indicated by the utterance data, and records the updated utterance probability level in association with the specific agent and the utterance content in the storage section.