Malware Training Dataset Selection Using File Creation Dates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current malware detection datasets are large and contain a substantial amount of irrelevant content samples, leading to long neural network training periods and potential errors in detection.
Innovation Solution
A method and apparatus that utilize file creation dates to select relevant files for a content detection dataset, using a machine learning algorithm to analyze metadata and content to determine the relevance of files, particularly focusing on recent creation dates for malware samples.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If large malware datasets are used for training, then the detection coverage is improved, but the training time increases substantially
Solution Approach 1:
The patent extracts and removes irrelevant malware samples from large datasets based on file creation date analysis. By identifying and eliminating outdated or less relevant samples (those with older creation dates), the system reduces dataset size and training time while preserving the most relevant malware characteristics for effective detection.
Solution Approach 2:
The patent changes the parameter of dataset composition by filtering samples based on file creation date thresholds. This parameter-based filtering transforms a static large dataset into a dynamic, time-aware dataset that prioritizes recent malware samples, thereby reducing training time while maintaining detection coverage.
2Reliability
If large malware datasets are used for training, then the detection coverage is improved, but the detection accuracy deteriorates due to irrelevant information
Solution Approach 1:
The patent extracts and removes irrelevant malware samples from large datasets based on file creation date analysis. By identifying and eliminating outdated or less relevant samples (those with older creation dates), the system reduces dataset size and training time while preserving the most relevant malware characteristics for effective detection.
Solution Approach 2:
The patent applies different quality criteria to different portions of the dataset by prioritizing recent samples over older ones. This local quality approach ensures that the most relevant, recently created malware samples are weighted more heavily in training, improving detection accuracy for current threat landscapes while maintaining broad coverage.
3Measurement precision
If recent malware samples are prioritized in the dataset, then the detection accuracy for new malware is improved, but the dataset size is reduced
Solution Approach 1:
The patent changes the parameter of dataset composition by filtering samples based on file creation date thresholds. This parameter-based filtering transforms a static large dataset into a dynamic, time-aware dataset that prioritizes recent malware samples, thereby reducing training time while maintaining detection coverage.
Data Source
AI summary
A method and apparatus for generating a content detection dataset using file creation dates. The method accesses a database comprising data files. The files are analyzed by a machine learning model to determine file creation dates. The creation dates are used to identify relevant content files. The most relevant files are included into a content detection dataset as content samples. The dataset may be used for training machine learning based content detectors.