Business Semantic Network Construction via Automated Terminology Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to efficiently detect and consolidate business terminology across different knowledge domains in a company, leading to inefficiencies and high costs in creating and maintaining a domain-oriented semantic network.
Innovation Solution
Utilizing business objects and search engines like SAP Enterprise Search or TREX to build a semantic network by extracting and organizing business terminology, with a terminology extractor and crawler that automatically identifies and updates related terms, reducing manual effort and errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual methods are used to detect and consolidate business terminology, then accuracy can be maintained, but implementation time and costs increase significantly
Solution Approach 1:
The system enables self-service by allowing the semantic network to automatically detect, extract, and consolidate business terminology from business objects without manual intervention. The terminology extractor automatically crawls business objects, identifies relevant terms, and builds the semantic network structure, eliminating the need for manual terminology detection while maintaining scalability.
Solution Approach 2:
Manual mechanical processes of terminology detection and consolidation are replaced with an automated computational system. The patent substitutes human analysts with a computer-implemented methodology that uses search engines, crawlers, and automated extraction algorithms to detect and organize business terminology from business objects.
2Productivity
If automated methods are used to build the semantic network, then implementation time is reduced, but complexity of the system increases
Solution Approach 1:
The system achieves universality by designing a multi-functional platform that can handle multiple tasks: crawling business objects, extracting terminology, building semantic networks, and updating existing networks. This single automated system replaces multiple separate manual processes, reducing overall system complexity despite the advanced capabilities provided.
Solution Approach 2:
The patent introduces intermediary components such as search engines and terminology extractors that mediate between business objects and the semantic network. These intermediaries simplify the overall system architecture by providing standardized interfaces and abstraction layers, making the automated process more manageable despite its complexity.
3Loss of information
If comprehensive business terminology is extracted from all business objects, then completeness of the semantic network improves, but processing resources and costs increase
Solution Approach 1:
The system applies partial action by extracting terminology selectively from business objects based on relevance criteria rather than processing every single business object uniformly. The crawler and extractor focus on extracting only the necessary terminology related to the specific domain and context, avoiding unnecessary processing of irrelevant data while maintaining completeness of the semantic network.
Solution Approach 2:
The patent segments the terminology extraction process into manageable components: crawling individual business objects, extracting terms from each object, and consolidating them into the semantic network. This segmentation allows the system to process large volumes of business objects in smaller batches, reducing memory requirements and enabling parallel processing to optimize resource utilization.
Data Source
AI summary
Pre-existing business objects (e.g., component parts of large business applications) may already define a hierarchy of related terms and include a search index created by a pre-existing search function. A semantic network including a plurality of semantic knowledge domains may be constructed automatically, based on the objects' initial terms structure and search index, and further modified by search results and related terms returned by the search function. This way, a customer specific semantic network may be constructed automatically from pre-existing software components and terms.