Timeseries DCF Capabilities to Predict Data Interactions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional data management paradigms are inadequate in predicting and supporting data access processes and interactions in heterogeneous distributed environments like a Data Confidence Fabric (DCF), making it difficult to gain insights into suspicious activities and data behavior, especially since DCFs operate primarily at the application layer and lack comprehensive control over data movement.
Innovation Solution
The implementation of a system that maps DCF information, such as confidence scores and annotations, onto a timeline to create a timeseries view, allowing for the identification, assessment, and prediction of data behavior as data transits through the DCF, including determining data paths, identifying unexpected gaps, and predicting future interactions using machine learning models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional data management paradigms are used in a Data Confidence Fabric, then the system structure remains simple and manageable, but the ability to predict and support data access processes and interactions is insufficient
Solution Approach 1:
The patent introduces a timeseries service as an intermediary component between the data confidence fabric and prediction requirements. This service ingests DCF information, maps it to a timeseries view, and provides prediction capabilities without fundamentally redesigning the entire DCF architecture, thus adding prediction reliability while controlling system complexity through a dedicated intermediary layer
Solution Approach 2:
The patent transforms DCF information by mapping it to a timeseries view, adding a temporal dimension to the data. This allows prediction of data access processes and interactions by analyzing trends over time, thereby improving prediction capability while maintaining the original DCF structure through a dimensional transformation approach
2Ease of operation
If the Data Confidence Fabric operates only at the application layer, then the implementation scope is limited and easier to manage, but it is difficult to gain insights about data behavior and suspicious activities
Solution Approach 1:
The patent implements a feedback mechanism where the timeseries service continuously ingests DCF information, analyzes it through timeseries mapping, and generates predictions about data behavior. This feedback loop enables the system to gain insights into suspicious activities and data patterns while maintaining operational simplicity through automated analysis processes
Solution Approach 2:
The patent performs preliminary mapping of DCF information to a timeseries view, organizing data in a structured temporal format before analysis. This preliminary action enables more effective detection of suspicious activities and data behavior patterns, thereby reducing information loss while keeping the operational framework simple and manageable
Data Source
AI summary
One example method is performed in connection with a data confidence fabric. This method includes generating information about data as the data transits a data confidence fabric, ingesting the information, mapping the ingested information to a timeline, evaluating the timeline, based on the evaluating, generating a recommendation for an action concerning the data, and implementing the action in the data confidence fabric. In this method, the evaluating includes determining if any unexpected time gaps occurred as the data transited the data confidence fabric.