Peide Zhu

selected publications

1. MMM

   MRHF: Multi-stage Retrieval and Hierarchical Fusion for Textbook Question Answering

   \textbfZhu, Peide, Zhen Wang, Manabu Okumura, and 1 more author

   In To appear at International Conference on Multimedia Modeling (MMM) , 2024
2. CHIIR

   On the Effects of Automatically Generated Adjunct Questions for Search as Learning

   \textbfZhu, Peide, Arthur Câmara, Nirmal Roy, and 2 more authors

   In To appear at Proceedings of the 2024 conference on human information interaction and retrieval (CHIIR) , 2024
3. MMM

   A New Benchmark and OCR-free Method for Document Image Topic Classification

   Zhen Wang,  \textbfZhu, Peide, Fuyang Yu, and 1 more author

   In To appear at International Conference on Multimedia Modeling (MMM) , 2024
4. PRICAI

   Improving Long Content Question Generation with Multi-level Passage Encoding

   Peide Zhu

   In Pacific Rim International Conference on Artificial Intelligence, 2021
5. ICTIR

   Evaluating BERT-based Rewards for Question Generation with Reinforcement Learning

   Peide Zhu, and Claudia Hauff

   2021
6. EDM

   MOOC-Rec: Instructional Video Clip Recommendation for MOOC Forum Questions

   Peide Zhu, Jie Yang, and Claudia Hauff

   In Proceedings of the 15th International Conference on Educational Data Mining, Jul 2022

   Abs PDF

   In this work, we address the information overload issue that learners in Massive Open Online Courses (MOOCs) face when attempting to close their knowledge gaps via the use of MOOC discussion forums. To this end, we investigate the recommendation of one-minute-resolution video clips given the textual similarity between captions and MOOC discussions. We first create a large-scale dataset from Khan Academy video transcripts and their forum discussions. We then investigate the effectiveness of applying pre-trained transformers-based neural retrieval models to make a ranked list of video clips for a forum discussion. The retrieval models are trained with supervised learning and distant supervision to effectively leverage the unlabeled data(over 80% of all data). Experimental results demonstrate that the proposed method is effective for this task, by outperforming baseline by 0.208 in terms of precision. To the best of our knowledge, this is the first systematic research applying pre-trained transformer-based ranking models on MOOC clip recommendation.

7. NAACL

   Unsupervised Domain Adaptation for Question Generation with DomainData Selection and Self-training

   Peide Zhu, and Claudia Hauff

   In Findings of the Association for Computational Linguistics: NAACL 2022, Jul 2022

   PDF Code
8. EMNLP

   Answer Quality Aware Aggregation for Extractive QA Crowdsourcing

   Peide Zhu, Zhen Wang, Claudia Hauff, and 2 more authors

   In Findings of the Association for Computational Linguistics: EMNLP 2022, Dec 2022

   Abs PDF Code

   Quality control is essential for creating extractive question answering (EQA) datasets via crowdsourcing. Aggregation across answers, i.e. word spans within passages annotated, by different crowd workers is one major focus for ensuring its quality. However, crowd workers cannot reach a consensus on a considerable portion of questions. We introduce a simple yet effective answer aggregation method that takes into account the relations among the answer, question, and context passage. We evaluate answer quality from both the view of question answering model to determine how confident the QA model is about each answer and the view of the answer verification model to determine whether the answer is correct. Then we compute aggregation scores with each answer’s quality and its contextual embedding produced by pre-trained language models. The experiments on a large real crowdsourced EQA dataset show that our framework outperforms baselines by around 16% on precision and effectively conduct answer aggregation for extractive QA task.