The process Linked Data creation involves flow in which data are initially identified and selected, and subsequently cured, mapped tosemantic representations, connected, and finally published. Each element of this flow defines a set of services that will be createdwithin the scope of the proposed project. The complete flow will be executed through the Web interface of the prototype, which will befed with data exposed by the producing entities and, after processing, they will be exploited by consuming entities, as illustrated in figure below.
We describe below the steps of performing the service:
- Data Identification and selection: a Web application will be available that will allow users to define data sources and select the attributes to be transformed into Linked Data. Data exported as spreadsheets, XML or SQL dumps can be loaded into the repository.
- Data Entry Forms: Web interfaces for data entry within a set of well-defined areas (such as public events, public services, business information, etc.) will be available.
- Clean, conversion and transformation: From the selection of attributes, a service will be available for curing data. A predefined set of quality filters may be used for the elimination of low-quality data. Then, operations of formatting transformation and conversion / definition of units of measurement can be performed, followed by a process of data classification.
- Mapping: The final data set should be mapped to new data model, visible in the data Web. Ontologies, also known as vocabularies within the context of data Web, are the model of representation of these data in the data Web. The prototype will contain a mapping service / suggestion of mappings between the original model and the final model. Users can use these entities in the internal repository of vocabularies to the prototype.
- Linking: Suggestions of links with resources (URIs) data Web are available, ensuring that the dataset is interconnected with other datasets in the Linked Data Cloud.
- Storage and Publication: Finally the data will be stored and published using the principles of the Data Web, from now getting visible and available for consumption on the Web