INTRODUCTION & PROJECT OBJECTIVES

We present the overall project introduction and objectives.

MEET OUR TEAM

We present our team and provide contact details.

VIEW OUR RESOURCES

We present our project deliverables including group papers, individual papers and individual codebases.

INDIVIDUAL SECTIONS

SEQUENCE-TO-SEQUENCE-BASED PARSER

This study investigates the accuracy of the BART transformer model in the context of a sequence-to-sequence parser that will parse English sentence into EDS semantic graphs.

GRAPH-BASED PARSER: NODE PREDICTION

This study compares the use of BiLSTMs and transformers for performing node prediction for semantic graphs constructed under the EDS framework.

GRAPH-BASED PARSER: EDGE PREDICTION

This study presents additional research into the effectiveness of the BERT Transformer and Maximum Entropy Loss within the development of the Edge Prediction Component of a Semantic Graph Parser.