OptiqueNLQF: A natural language query formulation system based on Semantic Technologies
Tomas Stein Sæbu · NORA - Norwegian Open Research Archives · 2015
This master project investigates the possibilities of having a natural language query formulation system where users will be able to specify an initial version of a query.Instead of confining users to pointing, clicking, scrolling and searching the menus for the right classes and attributes of a visual interface, we want to provide the opportunity to express their information requests in their own words.We want to create a system that is usable for a wider range of people, and where the technical background and knowledge about query generation no longer needs to be a requirement.Making database searches available for the average worker will increase the information flow, when everyone can retrieve the information they are interested in without having to ask specialists, non-technical staff are free to focus on their primary tasks and the information flow is undisrupted.The OptiqueNLQF-system analyzes an input information request and generates a SPARQL query.The prototype developed in this thesis is capable of handling simple natural language input phrases and will generate a functioning SPARQL query.Throughout this thesis we will discuss the design choices which have been made and the implementation of the system, as well as what possible improvement can be made.