The OS-EM Semantic Toolkit is a group of functions for managing and exploring data, and creating queries all wrapped up in an intuitive GUI. Currently, the toolkit reads XLSX files and processes them into a graph database. It can also create databases directly from RDF, Turtle (TTL), and N-Triples (NT) files.
The toolkit can be used in several different ways. Once you have a database, you can run queries against it using the syntax-highlighting Query Panel. You can create queries for future use via the Insight Manager. You can check the consistency of your data and look for errors with the Quality Checker. The most common use-case is using the Perspectives and Insights to visualize the data.
The database is a triplestore of your data. Nothing happens without a database. Luckily, OS-EM Semantic Toolkit can generate a triplestore from a variety of sources. Check out the data loading page for more information on creating a triplestore. If you already have a triplestore, the Toolkit can connect to a variety of popular engines.
The system requires every database it opens to have a defined “Base URI” that uniquely identifies the data set. It can also contain (a very small set of ) optional metadata to make the GUI easier to use. Currently, the metadata consists of:
Insights are essentially saved SPARQL queries. Unlike storing raw SPARQL, however, an Insight allows a user to associate a label, description, and output display with the query. So, a user can write the query, decide what to call it, and how the results should be visualized.
Even better, Insights can have Parameters associated with them. Parameters allow the user to define a generic query in the Insight, and then bind values to its variables during before execution. The system automatically figures out which bindings go with which variables. It’s also smart enough to figure out which variables are dependent on other variables, and act accordingly. So, for example, the user might have one Parameter that lets the user choose what type of concept to map, and then a dependent Parameter that allows them to select a given instance of that type.
Perspectives are for organization. They have a name and description, and contain zero or more Insights. Thus, the user can organize Insights arbitrarily. For example, in a database containing population data, one Perspective might contain Insights dealing with disease which another Perspective might have Insights dealing with political divisions.
Perspectives and Insights can be stored in the database just like any other data. However, the Toolkit supports private Perspectives and Insights that are separate from the database. Private Perspectives and Insights are available only to their author.
