Cross-Channel Message Conversion for Multi-Turn Digital Assistants
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robot-based chat systems face challenges in adapting to multiple interaction channels without significant modifications, leading to inconsistencies in information presentation and interaction strategies, and are limited by single-round text question-and-answer modes in handling complex scenarios.
Innovation Solution
A method and apparatus for message processing that converts interaction messages from various channels into a predetermined data structure, enabling consistent interaction and performing tasks based on these messages, while supporting multi-round chats to enhance user experience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a robot-based chat system is designed to work on a single interaction channel, then the system complexity is low and development is easy, but the system cannot adapt to multiple interaction channels without significant modifications and lacks consistency across channels
Solution Approach 1:
The patent introduces a message conversion module as an intermediary component between different interaction channels and the core processing system. This module receives messages from various channels (WeChat, SMS, email, etc.), converts them into a unified data structure, and forwards them to the task execution module. This intermediary layer enables multi-channel support without requiring modifications to the core system, resolving the contradiction between adaptability and complexity.
Solution Approach 2:
The system is segmented into distinct functional modules: a message conversion module for handling channel-specific protocols, a task execution module for processing unified messages, and a database module for storing interaction data. This segmentation allows each module to be developed and maintained independently, enabling the system to adapt to multiple channels while keeping overall complexity manageable.
2Adaptability or versatility
If the system uses single-round text question-and-answer mode, then the implementation is simple, but the system cannot effectively handle complex scenarios requiring multi-turn conversations
Solution Approach 1:
The system performs preliminary actions by maintaining user context information across multiple interaction rounds. The database module stores historical conversation data and user profiles, which are retrieved and integrated with current messages. This preliminary preparation of context information enables the system to handle complex multi-turn conversations without requiring complex real-time processing, balancing capability enhancement with manageable complexity.
3Reliability
If different interaction channels use their own data structures, then each channel can be optimized independently, but information presentation and interaction strategies become inconsistent across channels
Solution Approach 1:
The message conversion module enforces homogeneity by converting all channel-specific messages into a unified data structure with standardized fields and formats. This unified structure ensures consistent information presentation and interaction strategies across all channels. The conversion process maintains channel optimization benefits while achieving cross-channel consistency, resolving the contradiction between reliability and ease of development.
Data Source
AI summary
A method, an apparatus, a device and a storage medium for message processing are provided. The method includes: obtaining, from a target interaction channel of a plurality of interaction channels, a first interaction message from a user in a chat between the user and a digital assistant; converting, based on a target data structure corresponding to the target interaction channel, the first interaction message into a second interaction message with a predetermined data structure; and performing, based on the second interaction message, a task indicated by the first interaction message.