●Chapter 1 Evolutionary Algorithm based Ontology Schema-level Matching Technique
1.1 Preliminaries
1.1.1 Ontology, Ontology Matching, Ontology Alignment
1.1.2 Similarity Measure
1.2 Optimizing Ontology Alignments through Memetic Algorithm Using both MatchFmeasure and Unanimous Improvement Ratio
1.2.1 MatchFmeasure and Unanimous Improvement Ratio
1.2.2 MA Using MatchFmeasure and UIR
1.2.3 Experimental Results and Analysis
1.2.4 Conclusion and Future Work
1.3 Using Problem-speciˉc MOEA/D for Optimizing Ontology Alignments
1.3.1 Multi-Objective Ontology Matching Problem
1.3.2 MOEA/D for Optimizing Ontology Alignments
1.3.3 Experimental Results and Analysis
1.3.4 Conclusion and Future Work
Chapter 2 Evolutionary Algorithm based Ontology Instance-level Matching Technique
2.1 Using Memetic Algorithm for Instance Coreference Resolution
2.1.1 Similarity Measure for Instance Coreference Resolution
2.1.2 Memetic Algorithm for Instance Coreference Resolution
2.1.3 Experimental Results and Analysis
2.1.4 Conclusion and Future Work
2.2 Many-Objective Instance Matching in Linked Open Data
2.2.1 Many-Objective Instance Matching
2.2.2 NSGA-III based Many-Objective Instance Matching
2.2.3 Experimental Studies and Analysis
2.2.4 Conclusion and Future Work
Chapter 3 Improving the Performance of Evolutionary Algorithm based Ontology Matching Technique
3.1 An Alignment-Oriented Segmenting Approach for Optimizing Large Scale Ontology Alignments
3.1.1 The Framework of Segment-based Large Scale Ontology Matching Approach
3.1.2 Source Ontology Partition
3.1.3 Target Ontology Segment Determination
3.1.4 Ontology Segment Matching through the Hybrid Evolutionary Algorithm
3.1.5 Experimental Results and Analysis
3.1.6 Conclusion
3.2 E±cient Ontology Matching Using Meta-Model assisted NSGA-II
3.2.1 Error Ratio based Dynamic Alignment Candidates Selection Strategy
3.2.2 NSGA-II for Optimizing Ontology Alignment
3.2.3 Gaussian Random Field Model
3.2.4 Experimental Results and Analysis
3.2.5 Conclusion and Future Work
3.3 Using Compact Memetic Algorithm for Optimizing Ontology Alignment
3.3.1 Hybrid Population-based Incremental Learning Algorithm
3.3.2 Experimental Studies and Analysis
3.3.3 Conclusion and Future Work
Reference
內容簡介
本體描述了領域間的概念以及概念間的關繫,是解決語義網上數據異質問題的方案。但是由於人類的主觀性,同一個實體在不同本體中可能擁有不同的名稱和描述方式,使得本體間存在異質問題。給定兩個描述一繫列離散的實體(實體可能是概念、關繫和實例)的本體,確定這些本體間的關繫的過程稱為本體匹配,本體匹配可以有效地解決本體異質問題。當本體中的實體規模龐大的時候,本體匹配問題是一個復雜的(非線性問題且有很多局部很優解)和費時的(大規模問題)問題,因此近似的求解方法通常被用於確定本體匹配結果。源自這一觀點,進化算法成為了求解本體匹配問題的有效方法。本書首先為本體概念層和實例層構建了不同的單目標、多目標和眾目標模型,然後針對性地給出了各種進化算法(如混合進化算法,NSGA-II和MOEA/D)來求解這些模型。很後,還描述了各種提高基於進化算法的本體匹配技術性能的方法,如本體劃分算法、緊湊編碼方案、並行匹配模......