Trend Identification Mechanism for Wireless Data Mining
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data mining techniques for targeted web advertising require large and costly storage and management infrastructure, deterring service providers, especially in wireless communications networks, from implementing this revenue opportunity.
Innovation Solution
A mechanism that identifies trends in user data streams to provide customized information, such as targeted advertisements, without the need to store all user data records, reducing the complexity and cost of data storage and management infrastructure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional data mining techniques are used to store all user data records for targeted advertising, then advertising effectiveness is improved, but storage infrastructure cost and complexity increase significantly
Solution Approach 1:
The patent extracts only the essential trend information from the complete user data stream, discarding redundant detailed records. Instead of storing all user event records, the system stores condensed trend data that captures the essential patterns needed for targeted advertising, thereby reducing storage requirements while maintaining advertising effectiveness
Solution Approach 2:
The patent segments the continuous data stream into discrete trend identification cycles, processing data in manageable portions rather than storing everything at once. The system divides the data stream into time-based segments and identifies trends within each segment, reducing the immediate storage burden while maintaining comprehensive analysis capability
2Measurement precision
If all user data records are stored for data mining, then targeted advertisement accuracy is improved, but data management cost increases
Solution Approach 1:
The system extracts only the essential trend information from the complete user data stream, discarding redundant detailed records. Instead of storing all user event records, the system stores condensed trend data that captures the essential patterns needed for targeted advertising, thereby reducing storage requirements while maintaining advertising effectiveness
Solution Approach 2:
The patent changes the data representation from detailed event records to condensed trend parameters. By transforming the data from raw event-level details to aggregated trend parameters (such as frequency, timing patterns, and behavior sequences), the system reduces the quantity of data while preserving the information needed for accurate targeted advertising
3Productivity
If a large data storage infrastructure is implemented, then data mining capability is improved, but implementation cost increases
Solution Approach 1:
The patent extracts only the essential trend information from the complete user data stream, discarding redundant detailed records. Instead of storing all user event records, the system stores condensed trend data that captures the essential patterns needed for targeted advertising, thereby reducing storage requirements while maintaining advertising effectiveness
Solution Approach 2:
The system creates a simplified copy of the data stream that retains only the essential trend information needed for advertising decisions. This condensed copy serves as the basis for data mining operations, eliminating the need for expensive infrastructure to handle the complete original data stream while maintaining analytical capability
Data Source
AI summary
To provide customized information to the user, a wireless communications network node receives a stream of data associated with a user. A first trend associated with at least a first attribute in the stream of data is identified, and based on the identified first trend, customized information is sent for presentation to the user at a mobile station.