Web service discovery using combined bi-term topic model and WDAG similarity

• Published in: 2017 11th International Conference on Information & Communication Technology and System (ICTS) pp. 235 - 240
• Main Authors: Baskara, Andreyan Rizky, Sarno, Riyanarto
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.10.2017
• Subjects: BTM; Information and communication technology; Probability distribution; Resource management; Search engines; semantic; Semantics; structural; WDAG; web service discovery; Web services
• ISBN: 9781538628256; 1538628252
• ISSN: 2996-1378
• DOI: 10.1109/ICTS.2017.8265676

In recent years, many web services had been published by service providers. Finding similar web services to replace existing web services that a business actor owned has become a challenging task. This issue is identified as web service discovery problem. Two approaches to address this problem are measuring the semantic and structural similarity of web services. These approaches are performed by utilizing information in Web Service Definition Language document. This paper proposed a method which combined semantic and structural similarity of web services using Bi-term Topic Model (BTM) and WDAG similarity. In the proposed method, web service structure is modelled into Weighted Directed Acyclic Graph (WDAG). Then BTM is used to mine topic on the modelled WDAG. Jenson-Shannon divergence is used to calculate topic similarity and WDAG similarity is used to calculate the structure similarity of WDAG. The result of experiment shows that the proposed method is applicable for web service discovery with average precision 83.78% and average recall 91.79%.