The Sheet2RDF Semantic Turkey extension
Sheet2RDF is also available as Firefox extension and it works with Semantic Turkey extension.
On the Semantic Turkey site, we provided detailed instructions for the installation of Semantic Turkey (note on versions: Sheet2RDF 1.2.1 requires Semantic Turkey 0.12). For the impatients, we provide a self-contained explanation of the whole process, including the installation of Semantic Turkey.
To install a Firefox extension, it is sufficient to drop the related .xpi file on the browser window, accept the installation, and then restart the browser. Since version 43, Firefox prevents by default the installation of unsigned extensions (such as Semantic Turkey and Sheet2RDF). Hoewever, it is possible to disable that check, by setting to false the preference xpinstall.signatures.required in about:config (you should type the latter in the address bar).
The installed Firefox extensions rely on the services provided by a Semantic Turkey server. In this case, in order to use Sheet2RDF, you need also the complementing Sheet2RDF extension for the server.
Since we are assuming that you haven't already installed Sheet2RDF, we suggest to use the complete distribution of the server, including Semantic Turkey, Sheet2RDF and CODA. Alternatively, if you already have a Semantic Turkey Server, you can automatically download and install the required components, by placing the features file of Sheet2RDF onto the deploy directory of the server installation.
The server does not really require an installation, since it is sufficient to unpack it. Before starting the server, check that you have installed Java 7 (no less, and preferably no more) and that the enviroment variable JAVA_HOME points to its installation directory. To start the server, execute st_server_run.bat (under Windows) or st_server_run.sh (under Linux) placed in the subfolder bin.
Sheet2RDF on Firefox
Once the installation is complete, you can start playing with Semantic Turkey. The user interface of Sheet2RDF consists of a tab, which can be opened by clicking on the datasheet icon shown on the Semantic Turkey toolbar.
The UI is composed of three panels:
- the left panel (1) provides a preview of the datasheet content;
- the right panel (2) allows to managing the PEARL document;
- the bottom panel (3) shows the rdf triples generated by the system.
Now, let's see how Sheet2RDF works.
The input file (the spreadsheet to be processed) can be chosen through the select button (1). Its content is then imported into the system through the load button (2). Once the input file is loaded, a preview of the datasheet is shown into the left panel. In the screenshot below, the column headers are colored in red, because the user must supply a valid mapping to an RDF property.
The spreadsheet columns have to be mapped to properties of the target dataset. A dedicated dialog can be opened by clicking on the small icon near the column name, presenting a range of forms for specifying the mapping.
The primary information is the property to be bound to the column, which can be manually typed in a text field or automatically looked up on the underlying ontology, either by exploring its property tree (by clicking on the blue rectangular icon) or by searching properties whose labels match the column header (by clicking on the magnifier icon).
The second core form allows to the select a CODA converter, which transforms the cell values into RDF terms (either URIs or literals). See the tutorial page, for hints on how to choose a suitable converter. When a column mapping is complete, the column header is colored in green (or in blue, if it is the column associated with the subject).
If you're working on a SKOS or SKOS-XL project, you should choose a skos:ConceptScheme by opening the SKOS panel (1), selecting the schemes tab (2) and then selecting a pre-existing scheme or creating a new one (3).
Sheet2RDF triplifies the data by following the transformations indicated in the PEARL document. A basic PEARL document can be generated automatically by the system through the generate pearl button (1). In most of the cases, it may be necessary to edit the code, in order to replace a placeholder (
<pls_provide_...>, that is, part of the PEARL which necessarily need to be filled by the user) with a ground value, or just for customizing the transformation differently from the default choices suggested by the system. A PEARL document can be saved (2), exported (3) into the filesystem and then reloaded (4) in case of future reuse.
Clicking on generate triples button (1), the system generates the rdf triples and show them in the bottom panel. In this panel, the triples will be shown with an alternation of background color. The triples with the same color are generated from the same row of the datasheet. Clicking on a specific row of the triples panel, the source datasheet row will be highlighted as in the figure below.
Finally, after clicking on the add button (2), all the triples are added to the project.