Mobile Log Parsing via Random Sampling and Metadata Tables
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for parsing mobile device log data are inefficient and complex, with brute-force methods experiencing efficiency decline and pattern recognition methods being time-consuming and having low universality and accuracy.
Innovation Solution
A method involving random sampling of mobile device log data to construct metadata tables and data dictionaries, allowing for efficient parsing and extraction of relevant information using a stepwise methodology that incorporates pattern recognition techniques for improved accuracy and maintainability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If brute-force methods are used for parsing mobile device log data, then all data fields can be extracted, but parsing efficiency declines sharply with increase in keywords
Solution Approach 1:
The patent segments the parsing process into two phases: (1) sampling phase where a subset of log data is used to identify and extract keywords, and (2) parsing phase where the extracted keywords are applied to the full dataset. This segmentation allows efficient keyword extraction without requiring brute-force parsing of all data, thus maintaining high parsing efficiency while extracting a comprehensive set of keywords.
Solution Approach 2:
The patent performs preliminary action by sampling a subset of log data before full parsing. During this sampling phase, keywords are identified and extracted in advance. This preliminary keyword extraction eliminates the need for inefficient brute-force parsing of the entire dataset, thereby improving parsing efficiency while still capturing all relevant keywords.
2Adaptability or versatility
If pattern recognition methods are used for parsing mobile device log data, then parsing can be performed, but the methods are time-consuming and have low universality and accuracy
Solution Approach 1:
The patent implements self-service by automatically sampling log data and extracting keywords without requiring manual pattern definition. The system autonomously identifies relevant keywords from the sampled data and uses them for parsing the full dataset. This automated approach eliminates the time-consuming manual pattern recognition process while improving universality, as the same method can be applied to different log data types without reconfiguration.
3Adaptability or versatility
If conventional parsing methods are used, then data can be processed, but maintainability is poor requiring drastic code changes when data changes
Solution Approach 1:
The patent introduces dynamics by making the keyword extraction adaptive to data changes. Instead of using fixed parsing rules that require code modifications when data formats change, the system dynamically extracts keywords from sampled data and automatically adapts to new data types and formats. This dynamic approach significantly improves maintainability, as the system can handle data format changes without requiring drastic code modifications.
Data Source
AI summary
A computer-executable method, computer program product, and system for parsing a data log from a device, the computer-executable method comprising receiving the data log from the device, sampling the data log to create a sampled portion of data, wherein the sampled portion of data includes each attribute in the data log, constructing a metadata table based on the sampled portion of data, constructing data dictionaries based on the sampled portion of data, and parsing the data log using the metadata table and the data dictionaries to create a data field mapping table.