Πλοήγηση ανά Συγγραφέα "Stavropoulos. Petros"

A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω

Τώρα δείχνει 1 - 1 από 1

Biomedical question answering
(2020) Stavropoulos. Petros; Σταυρόπουλος, Πέτρος; Athens University of Economics and Business, Department of Informatics; Koutsopoulos, Iordanis; Papageorgiou, Haris; Androutsopoulos, Ion
Question Answering and Machine Reading Comprehension (MRC) are crucial and complextasks in the Field of Natural Language Processing (NLP). In this thesis, we first introduceBioMRC, a novel biomedical dataset for cloze-type Question Answering, based on previouswork of the BioRead dataset, implementing the same baselines and models for comparison.We then develop two new models based on the SciBert model from AllenAI for solvingthe task of BioMRC. We use these pre-trained models as a transfer learning approachfor the BioASQ Task 8B Phase B, in a modified architecture, to investigate whether ourdataset can be used for improving exact answer Question Answering tasks. In addition,we experiment with other BERT-based models for solving the BioASQ task, which usethe SpanBert and BioBert models, as well as the Text-to-Text Transfer Transformer (T5)model, a generative Transformer-based model, which achieved the best results for the task.Moreover, we create a cloze-type version of the BioASQ Task 8B Phase B factoid instancessubset, which is used to boost the T5’s results when pre-trained on the BioMRC dataset,but can also be used in future work for automatic transformation of question-answerinstances to cloze-type question instances. Lastly, we perform error analysis of our bestmodel for the BioASQ task for exact answers, where we point out the shortcomings of thetask evaluation measures and some mistakes, that could be fixed by the BioASQ organizers,as an improvement of the task.