Website User Estimation via Lossy Compressed Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating website user numbers are inefficient and imprecise, especially for large websites, due to excessive data storage and sampling inaccuracies, particularly in shorter intervals.
Innovation Solution
A method using lossy compression to store and analyze data about accessing entities, mapping entity data onto an array of bits, and estimating user numbers based on activated bits, allowing for efficient storage and combination of data across intervals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all access records are stored for analysis, then user estimation precision is improved, but data storage requirements increase excessively
Solution Approach 1:
The patent extracts only the essential information needed for user estimation (entity identifiers and timestamps) from complete access records, storing only this critical subset rather than full log data. This extraction approach maintains estimation precision while dramatically reducing storage requirements by eliminating redundant information.
Solution Approach 2:
Instead of storing complete records and then analyzing them, the patent inverts the approach by directly storing pre-processed entity identification data that is optimized for estimation purposes. The data structure is designed from the ground up for efficient user counting rather than being derived from complete access logs.
2Measurement precision
If sampling rate is increased to improve precision, then user estimation precision is improved, but processing time and computational resources increase
Solution Approach 1:
The patent changes the fundamental parameter of data representation from complete access records to compact entity identification markers. This parameter transformation allows for faster processing while maintaining precision, as the simplified data structure enables rapid counting and analysis operations.
Solution Approach 2:
The patent uses cookie-based identifiers as lightweight copies of user identity information, replacing the need to process and analyze complete access records. These identifier copies enable fast user distinction and counting without the computational overhead of analyzing full session data.
3Quantity of substance
If fixed sampling percentage is used before data collection, then data storage is reduced, but estimation precision deteriorates when event count is low
Solution Approach 1:
The patent implements a dynamic data collection approach where the amount of stored information adapts based on the specific access patterns and time intervals. Rather than applying a fixed sampling rate, the system stores entity identifiers in a manner that automatically adjusts to the volume and diversity of traffic, maintaining precision across varying event counts.
Solution Approach 2:
The patent performs preliminary entity identification and data structuring at the point of access, preparing the data in advance for efficient analysis. By pre-processing and storing only essential entity markers rather than complete records, the system ensures precision is maintained regardless of subsequent analysis timing or event volume.
4Measurement precision
If re-sampling is performed to adjust sample rate, then user estimation precision is improved, but system complexity and processing overhead increase
Solution Approach 1:
The patent implements a self-service data structure where the stored entity identifiers automatically enable accurate user estimation without requiring external re-sampling operations. The data is organized in advance to be directly analyzable, eliminating the need for complex re-sampling algorithms and reducing system complexity.
Data Source
AI summary
The invention relates to a method and system for estimating a number of users of a website. According to the method, it is repeatedly determined that the website is accessed by an entity and data dependent on the entity is determined and stored. The stored data is repeatedly compressed using a lossy compressing algorithm and based on the compressed data a number of users of the website is estimated.