User Clustering via Inverted Index for Resource Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing user clustering methods based on user behavior data require frequent queries and storage operations across large databases, leading to high computational resource usage and storage space occupation, with limited flexibility.
Innovation Solution
A method and apparatus for user clustering that acquire clustering conditions including time and event selecting criteria, determine target time periods, generate association data, and select target users by creating an association table, allowing for efficient selection of users based on monitored time periods and event conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If user clustering is performed by routine querying and storage logic for user behavior database once for each predetermined time period, then user clustering can be performed based on user behavior data, but it occupies a large amount of computing resources and storage space
Solution Approach 1:
The patent pre-processes user behavior data to extract feature data and generates user behavior sequences in advance. By preparing the data structure beforehand and establishing an inverted index for time periods, the system avoids the need for full database queries during each clustering operation, significantly reducing computational resource consumption while maintaining clustering accuracy.
Solution Approach 2:
The patent extracts only the necessary feature data from the complete user behavior database and stores it in a condensed format. By extracting and storing only essential features rather than the entire raw data, the system reduces storage space requirements and accelerates subsequent clustering operations without losing the accuracy needed for effective user segmentation.
2Measurement precision
If user clustering is performed by routine querying and storage logic for user behavior database once for each predetermined time period, then user clustering can be performed based on user behavior data, but it occupies a large amount of storage space
Solution Approach 1:
The patent extracts only the necessary feature data from the complete user behavior database and stores it in a condensed format. By extracting and storing only essential features rather than the entire raw data, the system reduces storage space requirements and accelerates subsequent clustering operations without losing the accuracy needed for effective user segmentation.
Solution Approach 2:
The patent pre-processes user behavior data to extract feature data and generates user behavior sequences in advance. By preparing the data structure beforehand and establishing an inverted index for time periods, the system avoids the need for full database queries during each clustering operation, significantly reducing computational resource consumption while maintaining clustering accuracy.
3Productivity
If user clustering is performed by routine querying and storage logic for each predetermined time period, then clustering results can be obtained, but it has poor flexibility
Solution Approach 1:
The patent implements a dynamic data structure where user behavior sequences are organized by time periods with flexible querying capabilities. The inverted index structure allows the system to adapt to different time range requirements and clustering conditions without reprocessing the entire database, enabling flexible adjustments to clustering parameters and time periods while maintaining efficient processing speeds.
Data Source
AI summary
The present disclosure provides a method of user clustering, and the method includes: acquiring a clustering condition for a predetermined user group, wherein the clustering condition includes a time selecting condition and an event selecting condition; determining at least one target time period for each user behavior data in a user behavior database based on the time selecting condition; determining association data indicating a relationship between the each user behavior data and each target time period based on the each user behavior data and the each target time period; and selecting target association data for a time period to be monitored based on the time period to be monitored and the event selecting condition, so as to determine a target user belonging to the predetermined user group according to the target association data. The present disclosure also provides an apparatus of user clustering, a computer device and a non-transitory medium.