Website Error Detection via Access Category Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing web commerce platforms face challenges in identifying the root cause of website errors due to insufficient data analysis, as errors are difficult to recreate and understand, especially when they occur frequently among logged-in users, which can be misleading given the high frequency of logged-in user visits.
Innovation Solution
The implementation of an experience analytics system that compares user device attributes and access categories to determine which attributes are correlated with website errors, using techniques such as chi-squared analysis and threshold comparisons to identify relevant access categories and values associated with errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If traditional sales data analysis is used to monitor website performance, then business decisions can be made based on product success metrics, but the entire story of website errors and user experience issues cannot be detected
Solution Approach 1:
The patent segments website monitoring into multiple independent data collection components: error logs from server-side applications, user session data from client-side browsers, and access category information. Each segment captures specific aspects of website performance, and when combined, they provide comprehensive error detection capability without losing critical information
Solution Approach 2:
The patent introduces an intermediary analytics system that sits between the website infrastructure and decision-makers. This intermediary collects, processes, and correlates data from multiple sources (server logs, client sessions, access categories) to transform raw data into actionable error detection insights, enabling both complete information capture and efficient decision-making
2Loss of information
If developers rely on user state information where errors occur, then some error context can be obtained, but sufficient information to understand and reproduce errors is not provided
Solution Approach 1:
The patent merges three distinct data sources into a unified error analysis framework: server-side error logs, client-side session data, and access category information. By combining these previously separate information streams, the system provides complete error context including technical error details, user interaction state, and accessing characteristics, enabling full error understanding and reproduction
Solution Approach 2:
The patent creates a universal error detection system that handles multiple types of errors across different access categories (mobile, desktop, tablet, various browsers) through a single integrated framework. This multi-functional approach captures comprehensive error context regardless of the accessing device or user state, making error detection and reproduction feasible across diverse scenarios
3Productivity
If error frequency is analyzed without considering access category distribution, then simple error tracking is possible, but misleading conclusions are drawn about error relevance
Solution Approach 1:
The patent changes the analytical parameters from simple error counts to normalized error frequencies that account for access category distribution. By introducing access category-specific baseline metrics and comparing actual error frequencies against these adjusted parameters, the system maintains efficient error tracking while eliminating misleading conclusions about error relevance to specific user groups
Data Source
AI summary
Systems and techniques may be used website error detection. An example technique may include identifying an error corresponding to a website, retrieving a first set of user sessions where the error occurred and a second set of user sessions where the error did not occur, and determining, for an access category, whether members of a set of values of the access category are correlated to the error, using the first set of user sessions and the second set of user sessions. The example technique may include comparing a characteristic of at least one member of the set of values that was determined to be correlated to the error to a threshold. The example technique may include displaying an indication of the error and an indication of the at least one member of the set of values.