Table 1 A comparison of LOD-ABOG with existing knowledge base approaches
From: Linked open data-based framework for automatic biomedical ontology generation
Modules | Approaches | |||
|---|---|---|---|---|
Harris et al. (2015) | Cahyani et al. (2017) | Qawasmeh et al. (2018) | Proposed Approach (LOD-ABOG) | |
Text processing | ||||
Methods | NLP | NLP | Manual | NLP |
Concept Extraction | ||||
Methods | Dictionary lookup, Statistical information | Dictionary lookup | Manual | UMLS Mapping, LOD |
Evaluation | Accuracy 60% (domain independence), 90% domain specific | Accuracy 72% (represent concepts and relations) | Not available | recall 81.13%, precision 45.29%, F-measure 58.12% |
Relation Extraction | ||||
Methods | Syntactic Patterns | Syntactic Patterns | LOD | Rule based, Syntactic Patterns, Semantic Enrichment, LOD, BSF |
Evaluation | Accuracy 31–67% | Accuracy 72% (represent concepts and relations) | Accuracy in range (15–50%) | Recall 63.82%, Precision 66.77%, F-measure 65.26% |
Type of extracted data | List of concepts, relations between them, and synonyms | List of concepts, and relations between them | List of classes, relations between them, and instances of these class | OWL Ontology |