📚 I graduated from Jiangnan University in Wuxi, Jiangsu Province with a bachelor's degree.
⚡ Research interests: natural language processing (NLP), active learning, generative adversarial networks (GANs) and time series data imputation.
🌱 Currently living in Cangzhou, Hebei Province, China.
Recent Publications
Baohang Zhou, Xiangrui Cai, Ying Zhang, Xiaojie Yuan. An End-to-End Progressive Multi-Task Learning Framework for Medical Named Entity Recognition and Normalization. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). pages 6214-6224. [Paper]
Baohang Zhou, Xiangrui Cai, Ying Zhang, Wenya Guo, Xiaojie Yuan. MTAAL: Multi-Task Adversarial Active Learning for Medical Named Entity Recognition and Normalization. In Proceddings of the 35th AAAI Conference on Artificial Intelligence (AAAI 2021). pages 14586-14593. [Paper]
Ying Zhang, Baohang Zhou, Xiangrui Cai, Wenya Guo, Xiaoke Ding, Xiaojie Yuan. Missing Value Imputation in Multivariate Time Series with End-to-End Generative Adversarial Networks. Inf. Sci., Vol. 551. 67-82. [Paper]
I am retraining and evaluating the repository to asses the results mentioned in the paper. However, there are #s appearing in the steps created in the result folder, in both the Normalized term considered in NEN and the original text under it. Any explanation for this, or further documentation for results interpretation?
I found the last ~90 rows of the Mesh entitybase have "None" values. And among those entities, the UIs that start with "D" actually exist in the previous UIs. Can you elaborate on it?