Data Fusion System Reducing Model Complexity via Knowledge Graph
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current enterprise resource planning (ERP) systems face inefficiencies and poor prediction accuracy due to the complexity and redundancy of historical business data used in prediction models.
Innovation Solution
A data fusion system and method that includes a processor and storage device with modules for feature extraction, model building, model disassembly, and data fusion. This system converts form data into feature data, builds a prediction model, disassembles it into triples data, and fuses this data into knowledge graph data to reduce redundancy and complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If historical business data and historical information are used to build prediction models, then prediction functionality is achieved, but data redundancy and model complexity increase
Solution Approach 1:
The patent segments the prediction model into multiple specialized sub-models, each handling specific types of business data or prediction tasks. This segmentation reduces the complexity of individual models while maintaining overall prediction accuracy through coordinated execution of multiple simpler models.
Solution Approach 2:
The patent extracts and removes redundant data elements from the historical business data before feeding them into prediction models. By identifying and eliminating duplicate or irrelevant information, the system reduces data redundancy while preserving the essential features needed for accurate predictions.
2Adaptability or versatility
If more historical business data is used for prediction, then prediction coverage is improved, but system operation efficiency decreases
Solution Approach 1:
The patent performs preliminary data processing operations including data cleaning, validation, and feature extraction before the main prediction execution. By preparing and pre-processing the historical business data in advance, the system improves prediction coverage with comprehensive data while maintaining operational efficiency during actual prediction tasks.
3Reliability
If prediction models process large amounts of historical data, then prediction completeness is improved, but computing resource consumption increases
Solution Approach 1:
The patent applies local quality optimization by processing different types of historical business data with specialized sub-models tailored to specific data characteristics. This approach ensures comprehensive prediction coverage for various data types while optimizing computing resource allocation by matching data complexity with appropriate model complexity.
Data Source
AI summary
A data fusion system and a data fusion method that may effectively reduce redundancy and complexity after data amplification. The data fusion system includes a storage device and a processor. The storage device is configured to store a feature extraction module, a model building module, a model disassembly module and a data fusion module. The processor is electrically connected to the storage device. The processor is configured to execute the feature extraction module, the model building module, the model disassembly module and the data fusion module, and is configured to receive multiple form data. The feature extraction module extracts multiple feature data from the form data. The model building module performs a preprocessing operation on the feature data, and builds a prediction model. The model disassembly module disassembles the prediction model into multiple triples data. The data fusion module fuses the triples data into knowledge graph data.