Pivot Ontology for Heterogeneous Drug Data Integration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current solutions fail to effectively integrate and dynamically analyze semantically heterogeneous drug data sources for generating recommendations in medical prescription and pharmacovigilance tools, which is crucial for managing drug interactions and adverse effects.
Innovation Solution
A device and method for generating a pivot drug database using a pivot ontology that structures and integrates data from multiple heterogeneous sources, creating a graph-based representation of drug-related data for analysis and recommendation purposes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If multiple heterogeneous drug data sources are integrated using traditional methods, then the quantity of drug-related data increases, but the complexity of data structuring and terminology alignment increases significantly
Solution Approach 1:
The patent introduces a pivot ontology as an intermediary layer between heterogeneous drug data sources and the target database. This pivot ontology acts as a mediator that translates and aligns diverse terminologies (ATC, RxNorm, SNOMED CT, ICD-10) into a unified structure, thereby integrating multiple data sources without directly complexifying the integration system. The pivot ontology includes classes such as Drug, Ingredient, AdverseEffect, and Disease, with standardized relationships that simplify the integration process.
Solution Approach 2:
The patent segments the drug data integration process into distinct modular components: extraction module, pivot ontology module, and target database module. Each data source is processed independently through the pivot ontology, which breaks down the complex integration task into manageable segments. This segmentation allows each module to be developed and maintained separately, reducing overall system complexity.
2Adaptability or versatility
If traditional data integration methods are used, then data from different sources can be collected, but the ability to dynamically analyze and reason about drug interactions is insufficient
Solution Approach 1:
The patent transforms drug data from static records into a dynamic ontology-based structure with defined parameters and relationships. By changing the representation parameters from simple data fields to ontology classes and relationships (e.g., Drug-hasAdverseEffect-Disease), the system gains enhanced analytical versatility while maintaining reliability through formal logical constraints that ensure consistent reasoning about drug interactions.
3Loss of information
If heterogeneous classification rules from different data sources are applied, then the diversity of drug information is preserved, but the difficulty of querying and analyzing data across sources increases
Solution Approach 1:
The pivot ontology is designed with universal classes and relationships that can represent diverse drug information from multiple sources through a common interface. The ontology structure (Drug, Ingredient, AdverseEffect, Disease classes with standardized relationships) serves multiple functions: it preserves source-specific information through properties while enabling unified querying across all data sources, thus reducing query difficulty without losing information diversity.
Data Source
AI summary
A device for generating a pivot drug database implemented in a computer system, the device includes an extraction unit configured for extracting the data from a set of elementary drug data sources, the elementary drug data sources storing drug-related data, each elementary data source being associated with a representation of the data; a structuring unit configured for structuring the extracted data by applying a pivot ontology to the extracted data, the pivot ontology defining classes derived from one or more ontologies of the drug and relationships between the classes, which provides structured data associated with a graph representing the relationships between the classes corresponding to the structured data; the device being configured for generating the pivot drug database according to the graph and the structured data, the pivot database storing the structured data. Applications: drug interaction analysis tools, tool for assisting medical prescription.