A Comprehensive Review of Classic and Modern Techniques for Ontology Matching

EasyChair Preprint no. 10331

Abstract

Ontology matching facilitates interoperability and semantic integration across heterogeneous knowledge bases. Over the years, numerous techniques have been developed to effectively tackle the challenge of aligning ontologies. This paper provides a comprehensive review of classic and modern techniques for ontology matching. We present an overview of the fundamental concepts and principles underlying ontology matching, followed by an in-depth analysis of traditional methods, such as linguistic-based, structure-based, and instance-based approaches. Subsequently, we delve into the recent advancements in ontology matching, including machine learning-based techniques, deep learning-based strategies, and hybrid methods that combine multiple algorithms. We compare these techniques based on critical metrics, such as precision, recall, and F-measure, and discuss their strengths, limitations, and applicability in real-world scenarios. Additionally, we highlight the impact of ontological characteristics, such as size, complexity, and heterogeneity, on the performance of different matching techniques. Furthermore, we explore the challenges and open research directions in ontology matching, such as handling semantic drift, scalability, and incorporating contextual information. Therefore, this paper aims to provide researchers and practitioners with a comprehensive understanding of classic and modern ontology matching techniques, paving the way for further advancements and improvements in this critical area of semantic integration.

Keyphrases: Ontology Alignment, ontology mapping, Ontology Matching

@Booklet{EasyChair:10331,
  author = {Rohan Shaan},
  title = {A Comprehensive Review of Classic and Modern Techniques for Ontology Matching},
  howpublished = {EasyChair Preprint no. 10331},

  year = {EasyChair, 2023}}