Log Analysis System Grouping Transactions by Start End Logs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional log analysis systems inaccurately detect failures in electronic devices like POS terminals due to irregular operations, which can result in false positives even when the device is operating normally, as the occurrence sequence of logs may vary.
Innovation Solution
A log analysis system that identifies transactions from device logs, groups them based on start and end logs, and creates a learning model defining the number of occurrences for each log type, allowing for accurate failure detection by inspecting transactions against this model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the occurrence sequence of logs is learned as a normal state, then failure detection can be performed, but false positives occur when irregular operations are performed
Solution Approach 1:
The patent changes the parameter being monitored from the occurrence sequence of logs to the number of occurrences of each log type. This allows the system to accommodate variations in log sequence caused by irregular operations while still detecting actual failures through abnormal occurrence counts.
Solution Approach 2:
The patent segments the log analysis by grouping logs with the same start and end logs into the same group, then analyzing the number of occurrences of each log type within those groups. This segmentation allows for more precise failure detection that accounts for operational variations.
2Device complexity
If a fixed log sequence model is used, then failure detection is simplified, but the system cannot adapt to irregular operations
Solution Approach 1:
The patent introduces a dynamic approach by creating multiple learning models for different groups of logs (grouped by start and end logs). This allows the system to adapt to various operational patterns while maintaining a structured analysis framework.
Solution Approach 2:
The patent creates a universal analysis framework that handles both regular and irregular operations through the grouping mechanism. The same analysis method (counting log occurrences) is applied universally across different log groups, making the system versatile.
Data Source
AI summary
Provided is a log analysis system including: an identifying unit that identifies transactions from logs output from a device; a grouping unit that categorizes the transactions having both the same log related to start and the same log related to end into the same group; a learning unit that creates a learning model that defines the number of occurrences on a log type basis in the transactions of the same group; and an inspection unit that inspects a transaction of an inspection target based on the learning model.