Statistical Access Model Assignment via Profile Similarity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Access control systems face challenges in detecting anomalous access events due to the time-consuming process of collecting sufficient data to build statistical access models, especially when new individuals or devices are introduced, leading to potential misclassification during this initial data collection phase.
Innovation Solution
The method involves collecting data from existing access profiles and determining common characteristics to assign a statistical access model, thereby reducing the time needed to build models and increasing reliability by leveraging existing data, allowing for immediate detection of anomalous events.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is collected from scratch for each new access profile, then the statistical access model can be built with sufficient accuracy, but the time required to build the model increases significantly
Solution Approach 1:
The system performs preliminary actions by collecting and storing access event data from multiple individuals in advance, organizing it into access profiles before it is needed. When a new individual's model is required, the pre-organized profiles serve as ready-to-use templates, eliminating the need to collect data from scratch and significantly reducing model building time while maintaining reliability through similarity-based selection.
Solution Approach 2:
The system creates copies of existing statistical access models by identifying access profiles with similar characteristics to the target individual and assigning their pre-built models. This copying approach allows immediate anomaly detection capability for new individuals without waiting for sufficient data collection, while the similarity criterion ensures the copied model maintains adequate reliability for the specific context.
2Adaptability or versatility
If a new employee is hired, then the access control system can accommodate new users, but sufficient data cannot be collected immediately to build an accurate statistical access model
Solution Approach 1:
The system introduces an intermediary approach by using access profiles of similar individuals as mediators between the new employee and the statistical access model. Instead of requiring direct data collection from the new employee, the system finds intermediate profiles with similar characteristics and uses their pre-built models, enabling immediate accommodation of new users with maintained detection accuracy.
Solution Approach 2:
The system copies statistical access models from similar existing employees to new employees, allowing immediate accommodation of new users. By selecting profiles with similar characteristics (same location, time period, role) and assigning their models to new employees, the system enables instant adaptability while preserving reliability through the similarity-based selection criterion.
3Measurement precision
If statistical access models are built from historical data of individual access events, then accurate anomaly detection can be achieved, but the process is time-consuming and delays detection capability
Solution Approach 1:
The system merges multiple access profiles that share common characteristics (same location, time period, and similar access patterns) into reusable model templates. By combining data from multiple similar individuals beforehand, the system creates pre-built statistical models that can be quickly assigned to new users, maintaining measurement precision through the similarity criterion while dramatically improving productivity by eliminating redundant data collection and model building processes.
Solution Approach 2:
The system performs preliminary actions by pre-collecting and pre-processing access event data from multiple individuals, organizing it into access profiles with built-in statistical models before these models are needed. This preliminary preparation enables rapid deployment of anomaly detection for new employees while maintaining accuracy, as the heavy computational work of model building has already been completed in advance for similar profiles.
Data Source
AI summary
One or more embodiments include collecting data associated with a first access profile, collecting data associated with a second access profile, determining whether the data associated with the first access profile has a particular number of characteristics in common with the data associated with the second access profile, assigning a statistical access model associated with the second access profile to the first access profile based on the particular number of characteristics that the data associated with first access profile has in common with the data associated with the second access profile, and detecting an anomalous access event based on the statistical access model.