Social Media Message Labeling for Customer-Agent Interaction Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Monitoring customer opinions on social media is challenging due to asynchronous and many-to-many interactions, and the high volume of opinions makes manual monitoring difficult for business organizations.
Innovation Solution
A system and method that includes a conversation interface, conversation database, conversation analysis server, and knowledge repository to automatically label messages from customer-agent interactions on social media, using preprocessing, dialogue act analysis, issue status analysis, and issue/response identification modules to standardize and categorize messages, enabling automatic issue and response identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual monitoring of social media messages is performed, then analysis accuracy can be maintained, but the workload and time consumption increase significantly due to high message volume
Solution Approach 1:
The system segments the message analysis process into multiple independent modules: preprocessing module for text standardization, dialogue act analysis module for intent classification, issue status analysis module for problem identification, and issue/response identification module for extracting specific issues and responses. This segmentation allows parallel processing of different aspects of message analysis, significantly reducing time consumption while maintaining comprehensive analysis accuracy.
Solution Approach 2:
The system introduces intermediate processing layers between raw messages and final analysis results. The preprocessing module acts as an intermediary to standardize text before analysis, and the dialogue act analysis module serves as an intermediary to classify message intents before issue identification. These intermediary modules reduce the complexity of direct manual analysis while preserving accuracy.
2Productivity
If automated analysis systems are implemented, then processing speed increases, but system complexity increases due to multiple analysis modules
Solution Approach 1:
The system divides the automated analysis into distinct modular components (preprocessing, dialogue act analysis, issue status analysis, issue/response identification), where each module performs a specific function. This segmentation enables independent optimization and maintenance of each component while achieving high overall processing speed through parallel execution of multiple modules.
Solution Approach 2:
The analysis server is designed as a universal platform that handles multiple types of social media messages and interaction patterns through a single integrated system. The modular architecture allows the same framework to process different message types (complaints, inquiries, feedback) using the same pipeline of analysis modules, reducing system complexity compared to having separate systems for each message type.
3Reliability
If comprehensive message analysis is performed to identify issues and responses, then customer care quality improves, but computational resources and processing time increase
Solution Approach 1:
The preprocessing module performs preliminary text standardization and feature extraction before the main analysis stages. By preparing the text data in advance (normalizing formats, extracting key features, removing noise), the subsequent dialogue act analysis and issue identification modules can operate more efficiently with pre-processed input, reducing their computational resource requirements while maintaining comprehensive analysis quality.
Solution Approach 2:
The system applies partial analysis to different message types based on their characteristics. Not all messages require the full analysis pipeline - some messages can be quickly routed or handled with simplified analysis. This selective application of comprehensive analysis reduces overall computational resource consumption while ensuring that messages requiring detailed issue and response identification receive the necessary comprehensive treatment.
Data Source
AI summary
A system, method and non-transitory computer readable medium for labeling a plurality of messages from a customer-agent interaction on a social media service to identify an issue and a response are disclosed. For example, the system includes a conversation interface, a conversation database coupled to the conversation interface, a conversation analysis server coupled to the conversation database and a conversation knowledge repository coupled to the conversation analysis server. The conversation analysis server includes a preprocessing module, a dialogue act analysis module, an issue status analysis module and an issue/response identification module.