Java Semantic Web & Linked Open Data webapps – Part 1
Target Functionality – what is my application and what parts of the semantic web do I want and/or need to deliver. Does the application want/need to be part of the Linked Open Data web ?
|Linked Open Data webapp||Semantic backed J2EE webapp|
|The application will provide a
This will be an application that uses a body of information from an historic text called the “Lewis Topographical Dictionary of Ireland” – a pre-famine “gazeteer” of all the civil parishes in Ireland in 1842 – and provides a means to semantically explore it, query it, and search it. When a location is found, its related locations, population, the members of parliament or “houses”, natural resources, and administrative description will all be presented alongside so that a one-stop page allows a user to follow whatever their interest is, within the confines of the text, and also reach out to encyclopaedia information (dbPedia) and location based features (geonames). This would allow someone to find places where the “Earl of Shannon” had an interest, plot on a map, what places bounded his location, whether he had more than one location to himself, what mutual friends he may have had, and by example, who was born in those locations before 1842 (dbpedia). The data behind this would also be made available via a public sparql endpoint so that other semantic applications on the web could make use of the information.
|The application will provide
This application explores a means to create a semantically backed application for a corporate entity. The idea being that by having application data stored in a semantic repository can be treated no differently than a traditional RDBMS backed application, provide the same functionality and fulfill fairly rigorous anonymity requirements, while also having the advantage of a semantic query capability for enhanced data mining and analytics. There isn’t a sparql endpoint, but there is a query capability, and this can be restricted by application or user. This means that unlike the Linked Open Data webapp, an authorization layer and cloaking mechanism for sensitive data must be added to the mix, as doing so at a sparql level isn’t possible – so some kind of object-sparql layer is required so that business and security logic can be applied to the sensitive data.