A service components pipeline model based on multi-source data extraction

Abstract

Different with the traditional scientific data (such as physics, materials, weather, etc.). The model focuses on the application data source and proposes a service components pipeline model based on multi-source data extaction. Some questions should be considered:firstly, the data modes are not uniformed and metadata representations are difficult to build uniformly, etc. In the service component publishing process, we build an open multi-source data extraction and service data package model for the transparent access to multi-source data service. Secondly, in order to avoid the potential bottlenecks and conflicts(structural conflicts and data collision) in various service components workflow, we dynamically predict the pipeline bottlenecks, analyze the potential structural conflict and data collision among service components and construct the specific-domain component pipeline service model by monitoring trends in data throughput capacity. In the verification process, the model takes ethnic minorities featured villages GIS service workflow (involving national architectural data, ecological data and national costumes data) as an application use case to further verify the methods effectiveness we proposed.