A5008960567- Topic Modeling
- Natural Language Processing Techniques
- Speech and dialogue systems
- Genetics, Bioinformatics, and Biomedical Research
- Genomics and Rare Diseases
- Multimodal Machine Learning Applications
- Biomedical and Engineering Education
- Domain Adaptation and Few-Shot Learning
Ewha Womans University
2020-2021
Automatic question generation (QG) serves a wide range of purposes, such as augmenting question-answering (QA) corpora, enhancing chatbot systems, and developing educational materials. Despite its importance, most existing datasets predominantly focus on English, resulting in considerable gap data availability for other languages. Cross-lingual transfer QG (XLT-QG) addresses this limitation by allowing models trained high-resource language to generate questions low-resource In paper, we...
Multi-task learning (MTL) approaches are actively used for various natural language processing (NLP) tasks. The Multi-Task Deep Neural Network (MT-DNN) has contributed significantly to improving the performance of understanding (NLU) However, one drawback is that confusion about representation tasks arises during training MT-DNN model. Inspired by internal-transfer weighting MTL in medical imaging, we introduce a Sequential and Intensive Weighted Language Modeling (SIWLM) scheme. SIWLM...
In response to the increasing use of interactive artificial intelligence, demand for capacity handle complex questions has increased. Multi-hop question generation aims generate that requires multi-step reasoning over several documents. Previous studies have predominantly utilized end-to-end models, wherein are decoded based on representation context However, these approaches lack ability explain process behind generated multi-hop questions. Additionally, rewriting approach, which...
Recent efforts have aimed to utilize multilingual pretrained language models (mPLMs) extend semantic parsing (SP) across multiple languages without requiring extensive annotations. However, achieving zero-shot cross-lingual transfer for SP remains challenging, leading a performance gap between source and target languages. In this study, we propose Cross-Lingual Back-Parsing (CBP), novel data augmentation methodology designed enhance SP. Leveraging the representation geometry of mPLMs, CBP...
This paper describes a community effort to improve earlier versions of the full-text corpus Genomics & Informatics by semi-automatically detecting and correcting PDF-to-text conversion errors optical character recognition during first hackathon Annotation Hackathon (GIAH) event. Extracting text from multi-column biomedical documents such as is known be notoriously difficult. The was piloted part coding competition ELTEC College Engineering at Ewha Womans University in order enable...
Conversational question--answer generation is a task that automatically generates large-scale conversational question answering dataset based on input passages. In this paper, we introduce novel framework extracts question-worthy phrases from passage and then corresponding questions considering previous conversations. particular, our revises the extracted answers after generating so exactly match paired questions. Experimental results show simple answer revision approach leads to significant...
Conversational question answering (CQA) facilitates an incremental and interactive understanding of a given context, but building CQA system is difficult for many domains due to the problem data scarcity. In this paper, we introduce novel method synthesize with various types, including open-ended, closed-ended, unanswerable questions. We design different generation flow each type effectively combine them in single, shared framework. Moreover, devise hierarchical answerability classification...