UIT students have a scientific paper published at the MAPR2023 International Conference.

Sun, 22/10/2023 - 23:44

Title: "ViQP: Dataset for Vietnamese Question Paraphrasing"

Authors:

Nguyen Quang Sang - 19522122 - CNCL2019 - Primary Author

Vo Dinh Thuc - 19522322 - CNCL2019 - Co-Author

Nguyen Phan Anh Duc - 19522573 - CNTT2019 - Co-Author

Tran Thanh Dang - 20520428 - CNCL2020 - Co-Author

Supervisor: Master Nguyen Van Kiet

Abstract:

Paraphrasing questions is a crucial task in natural language processing, which has various applications such as information retrieval, question answering, and dialogue systems. Despite the importance of this task, there has been little work on Vietnamese question paraphrasing. To fill this gap, we created a new dataset consisting of pairs of questions, each containing a source question and a paraphrased question. The dataset was constructed by collecting questions from various sources and then paraphrasing them using a team of human annotators. We evaluate the questions generated from the model with the BLEU measure and compare it with human performance. We assess the questions generated from the model with the BLEU measure and compare it with human performance and ChatGPT.

We express our gratitude to Master Nguyen Van Kiet, Associate Dean of the Faculty of Information Science and Technology, for his dedicated guidance and for pointing out our limitations during the research and publication of this international scientific paper.

The 6th International Conference on Multimedia Analysis and Pattern Recognition (MAPR), sponsored by the Vietnamese Association on Pattern Recognition (VAPR), was held in Quy Nhon on October 5-6, 2023. The conference aims to bring together researchers and experts from academia and industry to share the latest research findings, experimental results, and strengthen potential collaboration opportunities in pattern recognition, multimedia analysis, and related fields. Key topics of interest include Pattern Recognition and Machine Learning, Multimedia Analysis, Biomedical Image Analysis and Biometrics, Computer Vision and Robot Vision, Document Analysis and Recognition, and Applications.