Ontology Co-occurrence Network for Semantic Web Term Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The Semantic Web faces challenges in automatically mapping natural language or vocabulary words to appropriate ontology terms due to ambiguity and the need for manual knowledge of ontologies, making it difficult for users to efficiently describe, search, or query data without prior familiarity with ontologies.
Innovation Solution
A system and method that utilize an ontology co-occurrence network and term index to determine the most suitable ontology context and terms for a given set of input words by ranking term sets based on consistency and popularity, allowing users to input natural language or vocabulary words and providing ranked term sets for mapping without requiring prior knowledge of ontologies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual ontology selection is used, then ontology term accuracy is improved, but user expertise requirement increases and operation complexity increases
Solution Approach 1:
The system performs automatic ontology term mapping without requiring user expertise. The mapping service autonomously analyzes input words, queries the ontology co-occurrence network, and returns appropriate ontology terms, eliminating the need for manual ontology selection by users.
Solution Approach 2:
The patent introduces an intermediary mapping service that acts as a bridge between natural language input and ontology terms. This service uses the ontology co-occurrence network as an intermediate data structure to automatically determine the most appropriate ontology context and terms, resolving the contradiction between accuracy and ease of use.
2Measurement precision
If comprehensive ontology coverage is provided, then term matching accuracy is improved, but system complexity increases and computational expense increases
Solution Approach 1:
The patent segments the large ontology space into manageable units by creating an ontology co-occurrence network. This network organizes ontology terms into clusters based on co-occurrence relationships, allowing the system to efficiently search only relevant portions of the ontology space rather than examining all ontologies comprehensively.
Solution Approach 2:
The system performs preliminary processing by pre-computing the ontology co-occurrence network and storing it in a term index. This preliminary action enables faster querying and reduces computational expense during actual mapping operations, as the heavy lifting of ontology analysis is done in advance.
3Ease of operation
If automatic mapping is implemented, then ease of operation is improved, but handling of ambiguous words becomes difficult
Solution Approach 1:
The system uses feedback from the ontology co-occurrence network to resolve ambiguities. When multiple ontology contexts are possible for an input word, the network provides feedback on which contexts are most relevant based on co-occurrence statistics, allowing the system to automatically select the most appropriate ontology terms even for ambiguous words.
Data Source
AI summary
A system and method to map a set of words to a set of ontology terms, the method including determining a starting point for ontologies including terms matching a set of words, determining a term set corresponding to the set of words in an ontology context of each of the starting points, ranking the term sets determined for all of the starting points, and providing an output of the term sets in a ranked order.