File System Metadata Analytics Using Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional systems for measuring file similarity in file systems are often time- and resource-intensive, relying on examining file contents and are less effective due to unreliable user-applied metadata tags, making it burdensome to identify related files and users for policy enforcement.
Innovation Solution
A computer-implemented method using machine learning models to analyze metadata tokens within a file system, training data vectors representing frequently accessed files and files with common ancestors, to determine related files, users, and tags, enabling efficient correlation and action suggestions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional systems examine file contents to measure file similarity, then measurement precision is improved, but productivity deteriorates due to time- and resource-intensive processing
Solution Approach 1:
The patent extracts and utilizes file metadata (names, paths, timestamps, permissions) separately from file contents to perform similarity measurements. This extraction approach allows the system to achieve meaningful file similarity assessment without the computational burden of content examination, resolving the contradiction between measurement precision and productivity.
Solution Approach 2:
The patent segments file information into metadata components (name, path, timestamp, permissions) and content components. By focusing analysis on the metadata segment, the system achieves efficient similarity measurement that maintains practical accuracy while dramatically improving processing speed and resource utilization compared to full content analysis.
2Productivity
If user-applied metadata tags are used to find file similarity, then productivity is improved, but reliability deteriorates due to unreliability of user-applied tags
Solution Approach 1:
The patent introduces an intermediary automated tagging system that uses machine learning models trained on file metadata patterns. This intermediary processes file metadata objectively to generate tags, mediating between raw metadata and user needs while eliminating the reliability issues of manual user tagging.
Solution Approach 2:
The system implements self-service automated tagging where the file system itself generates meaningful tags through pattern recognition algorithms. Files automatically receive tags based on their metadata characteristics without requiring user intervention, ensuring consistent and reliable tagging across the entire file system.
3Reliability
If manual tagging of individual files is performed, then reliability of metadata tags is improved, but ease of operation deteriorates due to burdensome user activity
Solution Approach 1:
The system implements self-service automated tagging where the file system itself generates meaningful tags through pattern recognition algorithms. Files automatically receive tags based on their metadata characteristics without requiring user intervention, ensuring consistent and reliable tagging across the entire file system.
Solution Approach 2:
The patent introduces an intermediary automated tagging system that uses machine learning models trained on file metadata patterns. This intermediary processes file metadata objectively to generate tags, mediating between raw metadata and user needs while eliminating the reliability issues of manual user tagging.
4Productivity
If machine learning models are trained on file metadata tokens, then productivity is improved through efficient analytics, but device complexity increases due to model training and inference requirements
Solution Approach 1:
The patent applies preliminary action by pre-training machine learning models on historical file metadata during system setup or offline periods. Once trained, these models are deployed for rapid inference on new files, achieving high productivity during operation while concentrating the complexity burden in the preliminary training phase rather than during ongoing file operations.
Data Source
AI summary
The disclosed computer-implemented method for file system metadata analytics may include (i) creating a set of training data to train a machine learning model to analyze tokens that describe files within a file system, the set of training data comprising a first set of vectors, wherein each vector represents tokens that describes files that are frequently accessed by a common set of users, and a second set of vectors, wherein each vector represents tokens that describes files with common file path ancestors, (ii) training, using the set of training data, the machine learning model, (iii) determining, by providing at least one input token to the machine learning model, that the input token is related to at least one additional token, and (iv) performing an action responsive to observing the input token and involving the additional token and the file system. Various other methods, systems, and computer-readable media are also disclosed.