Deep Learning-Based Text Classification to Improve Web Service Discovery

• Autores: Hadj Madani Meghazi, Sid Ahmed Mostefaoui, Moustafa Maaskri, Youcef Aklouf
• Localización: Computación y Sistemas (CyS), ISSN 1405-5546, ISSN-e 2007-9737, Vol. 28, Nº. 2, 2024, págs. 529-542
• Idioma: inglés
• Enlaces
  • Texto completo
• Resumen

  • Abstract: Due to the rising number of firms and organizations offering access to their business data or resources on the internet through APIs, there has been a significant increase in the number of web APIs. This poses a difficulty in swiftly and effectively finding online APIs. In order to tackle this problem, the introduction of service classification has been implemented to streamline the process of finding services within a vast array of options. Prior approaches have endeavored to classify web services based on semantic characteristics, although their precision has been constrained. This work introduces a novel strategy named “DeepLAB-WSC” to improve the identification of web services. The approach specifically emphasizes actions derived from textual descriptions of web services and utilizes advanced techniques from deep learning-based text classification. The suggested methodology was evaluated using a real-world web API dataset and achieved superior results compared to existing state-of-the-art research.