In this dissertation, we present a novel approach for user-driven composition of Web APIs by adopting end-user development paradigm. Automated Web service or Web API composition driven by end-users requires dealing with three major research thrusts: (i) heterogeneity of the Web services due to different protocols they use, (ii) automated addition of semantic annotations in the service descriptions and (iii) involvement of all groups of end-users in the composition process by providing an easy and accessible environment according to their requirements. Nowadays, variety of services on the Web is increasing considering pervasive computing environments where different devices often use different technologies, which make the communication complex and sometimes even troublesome. Sensor Web has a central role in addressing new sources of information that need to be dynamically integrated into systems. Although, the HTTP protocol has been used as a universal mean for tunneling messages in business-to-business scenarios, interoperability between protocols such as WSDL/SOAP for Web Services or OGC/SOS/SPS for sensors is missing. We consider this issue and investigate the correct and complete use of the HTTP protocol to publish, manage, and operate services on the Web by fully exploiting the REST (REpresentational State Transfer) principles. We first propose an REST-based framework for providing a common way of interactions among heterogeneous Web Services. To accomplish this, we employ a common RESTful interface for services interactions by defining URI scheme to describe resources, the HTTP idioms (methods, headers, status codes) to interact with the service, the domain application protocol, and the media type for hypermedia-driven interactions. Although there are several description formats and tools available to describe REST Web APIs in human and machine readable formats, adding semantic annotation to these descriptions still requires manual work. To resolve this issue, we propose a solution to enrich the Web API descriptions by adding semantic annotations semi-automatically through API profiles that links properties to concepts in shared vocabularies. With this enriched descriptions, the compositions can be done either by directly linking the outputs and inputs of selected APIs, or by including transformation services that transform and make outputs compatible to inputs according to the semantic relations hold in the annotations. This work proposes an extension of the Open API Initiative (OAI) specification to create comprehensive descriptions and focuses on the emerging concept of API profiling to add descriptive information of data semantics by addressing Dublin Core Application Profile (DCAP) guidelines. One of our driving goals is to involve all groups of end-users in the Web API composition process. In the twenty first century, we are experiencing a rapid growth of elder people, who are using technologies for different purposes. Due to their different characteristics, the analysis of the context of elder domain has become a scientific challenge. Thus, we investigate the requirements, barriers and attitudes towards using systems or services of this group of people to understand the possibilities of transforming their roles from end-users to end-user developers. Moreover, we analyze different interface design guidelines to design a tool accessible by all groups of end-users including elders. We implement a prototype considering the underlying concepts and adopted design guidelines. Finally, we conduct an evaluation of the prototype using usability questionnaires.

(2017). User-Driven Composition of Web APIs: Bridging the Gap between Users’ Requirements and Technology Constraints. (Tesi di dottorato, Università degli Studi di Milano-Bicocca, 2017).

User-Driven Composition of Web APIs: Bridging the Gap between Users’ Requirements and Technology Constraints

LUCKY, MEHERUN NESA
2017

Abstract

In this dissertation, we present a novel approach for user-driven composition of Web APIs by adopting end-user development paradigm. Automated Web service or Web API composition driven by end-users requires dealing with three major research thrusts: (i) heterogeneity of the Web services due to different protocols they use, (ii) automated addition of semantic annotations in the service descriptions and (iii) involvement of all groups of end-users in the composition process by providing an easy and accessible environment according to their requirements. Nowadays, variety of services on the Web is increasing considering pervasive computing environments where different devices often use different technologies, which make the communication complex and sometimes even troublesome. Sensor Web has a central role in addressing new sources of information that need to be dynamically integrated into systems. Although, the HTTP protocol has been used as a universal mean for tunneling messages in business-to-business scenarios, interoperability between protocols such as WSDL/SOAP for Web Services or OGC/SOS/SPS for sensors is missing. We consider this issue and investigate the correct and complete use of the HTTP protocol to publish, manage, and operate services on the Web by fully exploiting the REST (REpresentational State Transfer) principles. We first propose an REST-based framework for providing a common way of interactions among heterogeneous Web Services. To accomplish this, we employ a common RESTful interface for services interactions by defining URI scheme to describe resources, the HTTP idioms (methods, headers, status codes) to interact with the service, the domain application protocol, and the media type for hypermedia-driven interactions. Although there are several description formats and tools available to describe REST Web APIs in human and machine readable formats, adding semantic annotation to these descriptions still requires manual work. To resolve this issue, we propose a solution to enrich the Web API descriptions by adding semantic annotations semi-automatically through API profiles that links properties to concepts in shared vocabularies. With this enriched descriptions, the compositions can be done either by directly linking the outputs and inputs of selected APIs, or by including transformation services that transform and make outputs compatible to inputs according to the semantic relations hold in the annotations. This work proposes an extension of the Open API Initiative (OAI) specification to create comprehensive descriptions and focuses on the emerging concept of API profiling to add descriptive information of data semantics by addressing Dublin Core Application Profile (DCAP) guidelines. One of our driving goals is to involve all groups of end-users in the Web API composition process. In the twenty first century, we are experiencing a rapid growth of elder people, who are using technologies for different purposes. Due to their different characteristics, the analysis of the context of elder domain has become a scientific challenge. Thus, we investigate the requirements, barriers and attitudes towards using systems or services of this group of people to understand the possibilities of transforming their roles from end-users to end-user developers. Moreover, we analyze different interface design guidelines to design a tool accessible by all groups of end-users including elders. We implement a prototype considering the underlying concepts and adopted design guidelines. Finally, we conduct an evaluation of the prototype using usability questionnaires.
BANDINI, STEFANIA
Web API; Composition; Semantic Description; End-User Development
INF/01 - INFORMATICA
English
Scuola di dottorato di Scienze
INFORMATICA - 22R
27
2013/2014
(2017). User-Driven Composition of Web APIs: Bridging the Gap between Users’ Requirements and Technology Constraints. (Tesi di dottorato, Università degli Studi di Milano-Bicocca, 2017).
File in questo prodotto:
File Dimensione Formato  
phd_unimib_761118.pdf

accesso aperto

Descrizione: Tesi dottorato
Tipologia di allegato: Doctoral thesis
Dimensione 2.05 MB
Formato Adobe PDF
2.05 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10281/141783
Citazioni
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
Social impact