Key points are not available for this paper at this time.
Service discovery and composition are crucial tasks in the development of Web services, it is designed to select the appropriate services for each task and ensure that services are called in the correct order. Unlike traditional methods, we aim to solve the service composition problem from a new perspective by developing a novel service composition using semantics based natural language descriptions. The proposed algorithm employs a combination of natural language processing and semantic-based techniques to extract the functional semantics of service datasets and understand the user context. By parsing the user’s request, the method extracts possible sub-queries contained within the request and identifies the appropriate combination of sub-queries. Leveraging the BM25 algorithm and the ESIM semantic matching model, the method identifies atomic services with high correlation with sub-queries and determines the composable atomic services set according to the defined composition constraints. Experimental results validate the efficacy of our method.
Liu et al. (Wed,) studied this question.