Object Classification via Audit Journal Interaction Patterns
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Classifying objects on a computer system is challenging due to the large number of files and objects, making it difficult for administrators to manage and secure data, especially in complex systems where objects are created by multiple users, and existing methods like file extension classification are inefficient and prone to errors.
Innovation Solution
An apparatus and method that classify objects by evaluating audit journals to create logical groups based on interaction patterns, allowing administrators to implement security policies and manage data through graphical constructs and user input, enabling automated classification without manual inspection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If objects are classified manually by examining each one individually, then classification accuracy improves, but time consumption and labor cost increase significantly
Solution Approach 1:
The patent replaces manual mechanical classification with an automated electronic system that uses audit journal analysis and machine learning algorithms to classify objects automatically, eliminating the need for human examination of each object while maintaining high classification accuracy
Solution Approach 2:
The patent introduces an intermediary classification system that uses audit journals and interaction pattern analysis as a mediator between raw objects and their classifications, automatically deriving classification criteria from system usage data rather than requiring manual inspection
2Ease of operation
If file extension classification is used, then classification process simplifies, but classification accuracy deteriorates due to common extensions between unrelated applications
Solution Approach 1:
The patent performs preliminary analysis of audit journals and interaction patterns before final classification, using pre-collected data about how objects are actually used in the system to inform and improve classification accuracy beyond simple file extension matching
Solution Approach 2:
The patent changes the classification parameters from simple file extensions to multi-dimensional interaction patterns including access frequency, user behavior, and system usage data, thereby improving classification accuracy while maintaining automated process simplicity
3Reliability
If comprehensive object classification is implemented to ensure security policy coverage, then security reliability improves, but system complexity increases
Solution Approach 1:
The patent implements a self-service classification system that automatically analyzes audit journals and interaction patterns to generate classifications without requiring administrator intervention, thereby ensuring comprehensive security coverage while minimizing the complexity burden on system operators
Solution Approach 2:
The patent uses feedback from audit journals and system interaction data to continuously refine and improve object classifications, ensuring comprehensive security policy coverage through data-driven adjustments rather than manual system complexity
Data Source
AI summary
A method, apparatus and program product for classifying objects on a computer system. An audit journal of interactions with objects is evaluated to identify a plurality of logical groups of objects. A graphical construct, such as a hierarchical tree, is created from the plurality of logical groups. And, in response to user input directed to the graphical construct, at least one object is classified based upon a logical group within which the object is arranged.