Event Information Push System Using ML Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network information push methods lack analysis and management, resulting in low information push efficiency, often pushing irrelevant information to users, which can lead to user dissatisfaction and channel blocking.
Innovation Solution
An event information push method that collects and clusters network messages based on correlation degrees, content, publish time, and user relevance, aggregating them into event information, and identifying user groups to push relevant information to users, using machine learning techniques and a social networking platform like WeChat.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If network information is pushed to all users without analysis, then information coverage is maximized, but information push efficiency deteriorates and user satisfaction decreases
Solution Approach 1:
The system performs preliminary clustering of network messages into event categories before pushing information to users. By pre-processing and organizing messages into structured event information with associated user groups, the system prepares targeted content in advance, improving push efficiency without requiring complex real-time analysis during the pushing phase.
2Productivity
If relevant information is identified and pushed to specific user groups, then information push efficiency improves, but system complexity increases due to message clustering and user group identification
Solution Approach 1:
The system segments network messages into different event categories based on correlation degrees, and identifies specific user groups for each event type. This segmentation allows the system to process and push information in manageable chunks rather than handling all messages uniformly, improving efficiency while keeping the complexity of each processing module relatively simple.
Solution Approach 2:
The system introduces event information as an intermediary structure between raw network messages and user notifications. This intermediary layer aggregates related messages, extracts key attributes, and matches them with appropriate user groups, simplifying the overall processing pipeline by providing a structured intermediate representation.
3Measurement precision
If machine learning classification is used to categorize messages, then information relevance improves, but processing time and computational resources increase
Solution Approach 1:
The system applies machine learning classification selectively to messages that require categorization into event information, rather than processing all messages with the same level of complexity. By focusing computational resources on messages that benefit from classification and using simpler methods for routine messages, the system achieves good categorization accuracy while minimizing overall processing time.
Data Source
AI summary
The present disclosure provides an event information push method performed at a computer system, the method including: collecting network messages from a social networking platform; performing aggregation on the network messages, to obtain corresponding event information; obtaining, for each piece of event information, one or more attributes from the corresponding category of network messages; obtaining, from an event information and user database, a user group corresponding to the one or more attributes of each piece of event information; and pushing each piece of event information to a user in the user group corresponding to the event information.