Assessment Task Detail and Instructions:

Max word count: 5000 (Report)

What to submit: PDF/Word Report
Objective:
To create a dataset using Semantic Data Technologies (i.e. a data model using an ontology, a
dataset using RDF/RDFS) and an application interface (UI) that runs SPARQL queries at the
back-end and executes results in human-readable format.

Concept / Aim (05%)

You can choose any idea for the application considering if the application area is a suitable
one for a semantic-based application? Are the aims both useful and achievable? e.g. an
application where users can search or filter some data using search controls (like text fields,
checkboxes and radio buttons).
Design (30%)
Use Protégé to create an Ontology for your dataset. The Ontology should have good range of
classes, properties and individuals. Explain the constraints and axioms and your process of
designing the Ontology Model.
Implementation (30%)
Write code for a user interface (ideally using Python) which takes Natural Language queries
(text), converts user query to a SPARQL query which runs on your dataset and displays
results. Marks for this section will be given based upon the Natural Language Processing
Techniques, connection to the dataset, results display and quality of code. The code should
be dynamic and the examples in the report should include at-least 5 SPARQL queries.
Evaluation and Use (25%)
The marks for this section will be given for the SPARQL queries. You should consider
providing five different queries to show the use of the semantic concepts for your chosen
application. Please provide queries with the screenshots for the results (either run on Protégé
or your SPARQL end-point). These queries may include the queries from the step above
(implementation).
Research Mapping (10%)
What future lines of research your application/ontology can adapt. It may be a critically
reflect on your application. Maybe compare with alternative approaches or discuss limitations
and future work. It requires an extensive literature review.