Rebecca Dridan
Research Interests
- Combining statistics and deep NLP
- Parser evaluation
- Parse ranking
- Cross-framework parser comparison
- Japanese, and in general, multilingual NLP
- Question answering
Current Research
Most of my current projects involve enhancing deep parsing with statistics, specifically using the PET HPSG parser. In October, 2011 I moved to the University of Oslo to take up a postdoctoral position on the WeSearch project. Previous to that, I was employed for two years on the OLE project (Online Linguistic Exploration: Deeper, Faster, Broader Language Documentation) at the University of Melbourne, where my main role was to produce large amounts of multilingual parsed text. Both project have involved looking at supertagging configurations that can prioritise robustness or efficiency. I'm also looking at using statistical information to improve parsing of under-resourced languages, through developing methods of training statistical models without manual annotation.
I have a few side projects involving parser evaluation, looking at questions regarding how deep parsers should be evaluated, what is common between different frameworks, and what these differences mean in practice.
Publications
- Andrew MacKinlay, Rebecca Dridan, Dan Flickinger, Stephan Oepen and Timothy Baldwin: Using External Treebanks to Filter Parse Forests for Parse Selection and Treebanking, in Proceedings of the Fifth International Joint Conference on Natural Language Processing (IJCNLP 2011), Chiang Mai, Thailand (2011) [pdf bib]
- Dridan, Rebecca and Oepen, Stephan: Parser Evaluation Using Elementary Dependency Matching, in Proceedings of the 12th International Conference on Parsing Technologies (2011) [pdf bib]
- Andrew MacKinlay, Rebecca Dridan, Dan Flickinger and Timothy Baldwin: Cross-Domain Effects on Parse Selection for Precision Grammars, Research on Language and Computation (2010) [pdf bib]
- Rebecca Dridan and Timothy Baldwin: "Unsupervised Parse Selection for HPSG", in Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing (EMNLP 2010), Cambridge, USA (2010) [pdf bib]
- Rebecca Dridan: "Using Lexical Statistics to Improve HPSG Parsing", PhD Thesis, Saarland University (2009) [pdf bib]
- Rebecca Dridan, Valia Kordoni and Jeremy Nicholson: "Enhancing Performance of Lexicalised Grammars", in Proceedings of ACL-08: HLT, Columbus, USA (2008) [pdf bib]
- Jeremy Nicholson, Valia Kordoni, Yi Zhang, Timothy Baldwin and Rebecca Dridan: "Evaluating and extending the coverage of HPSG grammars: A case study for German", in Proceedings of the Sixth International Conference on Language Resources and Evaluation (LREC 2008), Marrakech, Morocco (2008) [pdf bib]
- Rebecca Dridan and Timothy Baldwin: "What to classify and how: Experiments in question classification for Japanese", in Proceedings of the 10th Conference of the Pacific Association for Computational Linguistics (PACLING), Melbourne, Australia (2007) [pdf bib]
- Rebecca Dridan: "Using Minimal Recursion Semantics in Japanese Question Answering", Master's Thesis, The University of Melbourne (2007) [pdf bib]
- Rebecca Dridan and Francis Bond: "Sentence Comparison using Robust Minimal Recursion Semantics and an Ontology", in Proceedings of the Workshop on Linguistic Distances, Sydney, Australia (2006) [pdf bib]