Trend Analysis via State Tables Without Time Dimension
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional Business Intelligence systems lack the ability to perform agile ad hoc query and report creation, especially when analyzing historical trends without an explicitly defined time dimension, relying on outdated data and requiring expert knowledge for data warehouse design.
Innovation Solution
A computer-readable storage medium with executable instructions that reconstructs historical data views using audit logs from real-time data sources, constructing state tables and transition tables to derive trend data without a specified time dimension, supporting cumulative or non-cumulative data analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional ETL and data warehouse systems are used, then historical trend analysis can be performed, but the system requires an explicitly defined time dimension and delayed data availability
Solution Approach 1:
The system performs preliminary actions by continuously maintaining current state tables that are updated in real-time as data changes occur. These pre-computed state tables are ready for immediate query, eliminating the need to wait for batch ETL processes to complete. When a user requests historical data, the system can immediately retrieve and compare relevant state tables without data delay.
Solution Approach 2:
The patent introduces an intermediary mechanism - the state table - that captures the current state of data at any point in time. This state table acts as a mediator between the real-time data source and the analytical query, allowing historical analysis without requiring the data to be stored in traditional time-dimensional data warehouse structures.
2Measurement precision
If traditional data warehouse design is used, then structured historical analysis is possible, but expert knowledge is required and ad hoc query creation is not supported
Solution Approach 1:
The system implements dynamics by allowing the state tables to be flexibly created and modified based on user queries rather than requiring pre-defined rigid data warehouse schemas. When a user submits an ad hoc query, the system dynamically determines which data elements and time periods are relevant, creates or selects appropriate state tables, and performs the analysis. This makes the system adaptable to any analytical need without requiring expert knowledge of data warehouse design.
3Productivity
If real-time data sources are used without time dimension, then current data is available, but historical trend analysis cannot be performed
Solution Approach 1:
The system performs preliminary actions by maintaining a series of state tables, each representing the data state at a specific point in time. As real-time data changes occur, new state tables are created or existing ones are updated, preserving historical snapshots. This allows the system to provide both current fresh data and historical context simultaneously, enabling trend analysis without sacrificing data freshness.
Solution Approach 2:
The patent applies parameter changes by introducing temporal parameters to the state tables. Each state table is associated with a time stamp or time period, allowing the system to track how data parameters change over time. This transformation enables historical analysis of real-time data sources by adding the time dimension parameter without altering the underlying real-time data structure.
Data Source
AI summary
A computer readable storage medium includes executable instructions to analyze a query associated with a report to identify query clauses. The query is applied against a data source that does not have a specified time dimension. An optimized audit log is formed from an existing audit log to store data changes of interest at the data source and temporal values associated with the data changes, where each data change corresponds to a query clause. A state table is constructed to specify the state of each query clause. Records in the optimized audit log are processed by temporal value to produce updated state tables. Transition tables are derived from the updated state tables to associate temporal values with data changes at the data source that impact a result produced by the query. Trend data is then supplied from the transition tables.