A Model-Driven Approach for Semantic Data-as-a-Service Generation

EasyChair Preprint 3489

Abstract

Nowadays, with the increasing number of data sources, especially in environmental domain, earth observation programs face major challenges for environmental data exploitation, mainly due to data sources heterogeneity of different types such as access techniques, used protocols, languages, data formats, etc. Although typical solutions abstract from this heterogeneity with a layer of data services, the development of such systems remains tedious in this context. In this paper, we propose an approach based on Model-Driven Engineering (MDE) combined with semantic annotations, to automate data service development on top of data sources. Our work contributes to the development of integrated service-based architectures driven by automatic service generation, data integration from existing environmental systems and automatic service annotations. Our solution, applied to the detection of natural disasters, provides 1) appropriate modelling of data sources and services to apply model-to-text (M2T) transformations, 2) automatic generation of Representational State Transfer (REST) data service code template, 3) automatic generation of semantically annotated Hypermedia-based descriptors of these services. We have implemented and evaluated our solution with a set of real datasources provided by the Sahara and Sahel Observatory (OSS), OpenWeatherMap and CHIRPS.

Keyphrases: Hypermedia Driven APIs, Model-Driven Engineering, Semantic RESTful Services

@booklet{EasyChair:3489,
  author    = {Hela Taktak and Khouloud Boukadi and Michael Mrissa and Chirine Ghedira-Guégan and Faiez Gargouri},
  title     = {A Model-Driven Approach for Semantic Data-as-a-Service Generation},
  howpublished = {EasyChair Preprint 3489},
  year      = {EasyChair, 2020}}