In May 2012, the first PATHS prototype and Web Service API was released. The system was subject to intensive testing in the period up to September 2012. During the testing period, feedback was gathered through combination of user trials, demonstrations and professional assessments. Specifications for the final prototype were finalized during the last quarter of 2012 and the development of the final prototype described in this document took place from January to April 2013.
The PATHS system includes hardware, software, applications and interfaces. Details about the development process itself a comprehensive data model and API and other documentation are available.
The data model is implemented as a combination of SQL and non-SQL databases and indexes combining the power of the PostgreSQL RDBMS with that of Apache Solr inverted index and Virtuoso RDF triple store. All relationships and references in the data model are implemented using persistent URIs as foreign keys. This allows for flexible integration between the three data stores.
The web service API is implemented on top of a wide range of server components and controls all data I/O operations towards the data layer. It consists of more than 30 different web methods grouped into seven Web Services. The Web Services communicate over the HttpGet, HttpPost, Soap and Soap 1.2 protocols. The default return format is JSON but the services are also capable of delivering XML.
The web application is implemented with a number of sophisticated end-user interfaces that rely and data received through Web Service requests for their operation. The application is user centric and emphasizes good interaction design as well as innovative modes of exploration. The prototype is designed to demonstrate the core functionality of the system and the potential of the navigation, information retrieval and content enrichment methodology proposed by the project.