Chatbot FAQ Registration via Poorly Resolved Question Prioritization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional FAQ registration processes for chatbots are laborious and time-consuming, leading to decreased user satisfaction due to inefficiencies in presenting suitable answers for input questions, especially when question data does not match stored data.
Innovation Solution
A response processing method that identifies and prioritizes poorly resolved question data for which answers were not identified or negatively evaluated, allowing for additional registration of answer data, thereby facilitating efficient registration and increasing the rate of suitable answer presentation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional FAQ registration processes are used for chatbots, then question data can be registered in the database, but the process becomes laborious and time-consuming
Solution Approach 1:
The system automatically identifies poorly resolved question data by analyzing chat logs and user interactions, then prioritizes and suggests answers for registration without requiring manual review of every question. The chatbot system serves itself by automatically generating candidate questions and answers based on actual usage patterns.
Solution Approach 2:
The system performs preliminary analysis of chat logs to identify poorly resolved questions before the FAQ registration process begins. By pre-processing and categorizing questions based on resolution status and frequency, the system prepares candidate FAQs in advance, reducing the time required for actual registration.
2Adaptability or versatility
If question data does not match stored data in the FAQ database, then the chatbot can handle new queries, but user satisfaction decreases due to inability to present suitable answers
Solution Approach 1:
The system incorporates feedback mechanisms where user interactions, chat logs, and resolution outcomes are continuously analyzed. Poorly resolved questions are identified based on feedback from actual chat sessions, and the system prioritizes registering answers for these questions to improve future response accuracy and reliability.
3Quantity of substance
If all question data is registered equally in the FAQ database, then comprehensive coverage is achieved, but registration becomes inefficient and time-consuming
Solution Approach 1:
Instead of treating all question data uniformly, the system applies different registration priorities based on local characteristics of each question. Questions are categorized by resolution status, frequency of occurrence, and difficulty level, with poorly resolved questions receiving higher priority for registration to optimize the balance between coverage and efficiency.
Data Source
AI summary
A non-transitory, computer-readable recording medium stores therein a response processing program that causes a computer to present, based on question data and information associating answer data, the answer data related to input question data. The response processing program causes the computer to execute a process including referring to a storage storing therein a response result for past question data input in the past, and displaying poorly resolved question data for which a related answer is not identified or a presented answer has been negatively evaluated; and setting specified poorly resolved question data among the displayed poorly resolved question data, as object question data for which information associating answer data is additionally registered. The displayed poorly resolved question data is displayed as priority question data having higher priority a greater is a poor resolution occurrence number thereof.