no code implementations • CCL 2021 • Qian Chen, Xiaoying Gao, Suge Wang, Xin Guo
“知识图谱问题生成任务是从给定的知识图谱中生成与其相关的问题。目前, 知识图谱问题生成模型主要使用基于RNN或Transformer对知识图谱子图进行编码, 但这种方式丢失了显式的图结构化信息, 在解码器中忽视了局部信息对节点的重要性。本文提出迭代信息传递图编码器来编码子图, 获取子图显式的图结构化信息, 此外, 我们还使用滑动窗口注意力机制提高RNN解码器, 提升子图局部信息对节点的重要度。从WQ和PQ数据集上的实验结果看, 我们提出的模型比KTG模型在BLEU4指标上分别高出2. 16和15. 44, 证明了该模型的有效性。”
no code implementations • EMNLP (ALW) 2020 • Kosisochukwu Madukwe, Xiaoying Gao, Bing Xue
Recently, a few studies have discussed the limitations of datasets collected for the task of detecting hate speech from different viewpoints.
no code implementations • 12 Dec 2023 • Trevor Londt, Xiaoying Gao, Peter Andreae, Yi Mei
This paper introduces a new CE representation and algorithm capable of evolving novel multi-path CNN architectures of varying depth, width, and complexity for image and text classification tasks.
no code implementations • 3 Dec 2020 • Trevor Londt, Xiaoying Gao, Peter Andreae
Results indicate that the algorithm evolves performant models for both datasets that outperform two of the state-of-the-art models in terms of model accuracy and three of the state-of-the-art models in terms of parameter size.
no code implementations • 3 Dec 2020 • Trevor Londt, Xiaoying Gao, Bing Xue, Peter Andreae
Researchers have not applied EDL techniques to search the architecture space of char-CNNs for text classification tasks.