The Concept

The Concept

While semantic software-developers rely on tools for engineering and running their applications (like TopBraid Composer and Protege), they do not have a common solution that combines semantics tools and supports generic application development, operation, and deployment. In addition to this integration challenge, there is no existing solution for knowledge acquisition by subject matter experts. No ontology discovery, search, and visualization tools have achieved any significant traction. A great integrated framework for semantic application development should offer moderate access to data models by clients, support users at the initial engineering stages and provide manipulation and visualization tools that allow people who are not versed in ontologies to discover the relationships and contents between models, augment models, and smoothly navigate the model space–therefore enabling continuous “in-field” improvement and maintenance of semantic applications.

Visualizing knowledge in a tangible and easily accessible manner is advantageous for many reasons (and not only in the complex calculation of credit scores) such as the semantic authentication of ontologies by subject matter experts and for helping those who are not experts in ontologies to extend their vocabularies. The Open Semantic Framework facilitates knowledge visualization through specialized, visualization-specific Knowledge packs with queries that allow client applications to discover knowledge models. For example, given a certain semantic node, the queries communicate that node’s properties, links to adjacent nodes, and information about its type. These queries are crucial since they can be used to obtain knowledge from our systems that could undermine economic motivations to contribute to it. Therefore, accessing generic exploration queries is strictly controlled. Using the Visualization Knowledge pack, visualizers assume the form of Open Semantic Framework client applications that use these queries to explore loaded ontologies.

Traditionally, the Open Semantic Framework only used the Open Database Connectivity platform (ODBC) to interact with Virtuoso and made use of iODBC drivers. The use of iODBC drivers in Ubuntu made using the system very complex because you had to recompile a significant amount of PHP packages to access that other Open Database Connectivity drivers.

The Open Semantic Framework version 3.2 refactored the code to allow a person to query all SPARQL HTTP endpoints. The improvement made in OSF’s version 3.2 was to allow users to access any triple store where the SPARQL HTTP endpoint is compatible with Virtuoso as well as the Open Semantic Framework.

The OSF version 3.3 made both these options possible. However one had to ensure that Virtuoso’s latest version was working well with the unixODBC drivers, which Ubuntu ships by default. This allowed people to use the ODBC platform through unixODBC drivers. The enhancement made it much easier to maintain Ubuntu/OSF because no packages were on hold, and the PHP packages could be updated at any time without the need to recompile them using the iODBC drivers. This major OSF release changed how web services interact with triple stores.

As mentioned earlier, OSF web services originally used an ODBC platform to interact with virtuoso. The only change made in the new OSF versions was incorporating SPARQL HTTP endpoints of the Virtuoso to send queries to it. This was a major improvement to the original version. The use of ODBC as the fundamental interaction platform between the triple store and the OSF had been associated with a lot of complexities. This made it necessary to switch to HTTP because the original version of OSF was not compatible with the Ubuntu-shipped PHP5 packages, and a person had to maintain their iODBC compatible packages. A major setback of this was that system administrators could not upgrade their Ubuntu system normally because PHP5 required to be upgraded using specific packages designed for that purpose.