Entity linking and API resource-based matchmaking for Web APIs composition

Web APIs composition is still a challenging task. In the classical semantic composition approach, the matchmaking process uses ontologies to ensure semantic matching between inputs and outputs. In this paper, we address the composition of Web APIs based on entity linking of Web APIs to a given knowl...

Full description

Saved in:
Bibliographic Details
Published in:Service oriented computing and applications Vol. 17; no. 2; pp. 93 - 108
Main Authors: Boustil, Amel, Tabet, Youcef
Format: Journal Article
Language:English
Published: London Springer London 01.06.2023
Springer Nature B.V
Subjects:
Application programming interface
Composition
Computer Appl. in Administrative Data Processing
Computer Science
Computer Systems Organization and Communication Networks
Datasets
Documentation
e-Commerce/e-business
Graphs
IT in Business
Knowledge
Linked Data
Management of Computing and Information Systems
Ontology
Original Research Paper
Semantic web
Semantics
Software Engineering/Programming and Operating Systems
OAS
Entity linking
Web API
Composition
Matchmaking
ISSN:1863-2386, 1863-2394
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Web APIs composition is still a challenging task. In the classical semantic composition approach, the matchmaking process uses ontologies to ensure semantic matching between inputs and outputs. In this paper, we address the composition of Web APIs based on entity linking of Web APIs to a given knowledge base (KB). We first consider the APIs resources that make up the main core of REST Web APIs. Then, we link both the API resource described in the OAS (Open API Specification) documents and the input/output attributes of Web API operations to entities of the given KB. Finally, we perform discovery and composition of Web APIs based on their related entities extracted from the KB. Our dependency graph realizes matchmaking based on identifying similar entities of the API resources and the attributes of the requested input and output parameters. Focusing on API resources reduces the space of searching for candidate operations. Moreover, extending the user query by related similar entities adds semantics to the matchmaking process in the KB context. Experiments on the developed prototype show the feasibility and the scalability of our proposed composition approach.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1863-2386
1863-2394
DOI:10.1007/s11761-022-00353-5