Answering Engine for Sport Statistics: Question Processing

Abstract

In recent years, there has been an increasing growth of interest among computer scientists for the topic of Linked Data and the Semantic Web. By connecting and publishing structured data from multiple sources, the Web enables us to retrieve specific information without needing to go through documents of unstructured text. Question answering systems can utilise the benefit of Linked Data, and enable users to ask question in a natural language in order to provide direct answers. In this thesis we implement a system that can answer natural language questions related to the field of Formula 1 statistics. We show how data is collected and connected based on a conceptual model, and go through the necessary steps for converting a question into a machine-readable query. We perform an evaluation of the system, both on component level and on the system as a whole. We analyse and discuss challenges and topics for improvements, before we conclude our work and summarise the most important steps to consider for future work.