Conversation Routing System for Bot Discovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current conversational bots are difficult to discover and often falter when asked out-of-domain questions, leading to missed opportunities for cross-promotion and user dissatisfaction due to the challenge of handling all types of user intents equally well.
Innovation Solution
A system that analyzes user queries and bot responses to determine the validity of replies, redirecting conversations to more appropriate bots based on a trained machine learning model, allowing for organic bot discovery and promotion without requiring developers to cover comprehensive out-of-domain coverage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a bot is developed to answer questions in a specific domain, then the bot can provide accurate responses within its domain, but the bot will falter when asked out of domain questions
Solution Approach 1:
The system segments the bot ecosystem into multiple domain-specific bots rather than using a single general-purpose bot. Each bot is specialized in a particular domain, allowing high accuracy within that domain while the overall system handles diverse queries through bot switching.
Solution Approach 2:
An intermediary system (the conversation routing mechanism) is introduced to determine which bot should handle a given query. This intermediary analyzes the query and routes it to the appropriate domain-specific bot, enabling the system to handle out-of-domain questions by transferring to a bot with relevant expertise.
2Adaptability or versatility
If more domain-specific bots are launched, then the system can handle more diverse user intents, but it becomes difficult for users to discover an appropriate bot
Solution Approach 1:
The system implements feedback mechanisms where user interactions with bots are analyzed to determine query validity and bot performance. This feedback is used to improve bot recommendations and routing, making bot discovery more effective over time as the system learns from actual usage patterns.
Solution Approach 2:
The system performs preliminary analysis of user queries to determine the appropriate bot before the user has to manually search for one. By pre-analyzing the query intent and routing to the correct bot automatically, the system eliminates the need for users to discover bots manually.
3Adaptability or versatility
If a bot handles all types of intents, then the bot can serve diverse users, but making a bot that can handle all types of intents equally well is very challenging
Solution Approach 1:
Instead of creating one bot that handles all intents, the system segments functionality across multiple specialized bots. Each bot maintains high reliability within its specific domain while the system as a whole achieves broad intent handling capability through the collection of specialized bots.
Data Source
AI summary
The present teaching relates to facilitating a user-machine conversation. In one example, a query is obtained from a user. The query is directed to a first conversational bot. A reply in response to the query is obtained from the first conversational bot. A degree of validity of the reply is determined based on the reply and the query. A second conversational bot is determined based on the query and the degree of validity. The conversation is directed to the second conversational bot with the query.