no code implementations • CCL 2020 • Hanzhong Qin, Chongchong Yu, Weijie Jiang, Xia Zhao
针对目前检索式多轮对话深度注意力机制模型DAM(Deep Attention Matching Network)候选回复细节不匹配和语义混淆的问题, 本文提出基于多头注意力和双向长短时记忆网络(BiLSTM)改进DAM模型的中文问答匹配方法, 该方法采用多头注意力机制, 使模型有能力建模较长的多轮对话, 更好的处理目标回复与上下文的匹配关系。此外, 本文在特征融合过程中采用BiLSTM模型, 通过捕获多轮对话中的序列依赖关系, 进一步提升选择目标候选回复的准确率。本文在豆瓣和电商两个开放数据集上进行实验, 实验性能均优于DAM基线模型, R10@1指标在含有词向量增强的情况下提升了1. 5%。
no code implementations • 14 May 2021 • Weijie Jiang, Zachary A. Pardos
With equity of educational outcome as the aim, we trial strategies for boosting predictive performance on historically underserved groups and find success in sampling those groups in inverse proportion to their historic outcomes.
no code implementations • 2 Jul 2019 • Zachary A. Pardos, Weijie Jiang
Collaborative filtering based algorithms, including Recurrent Neural Networks (RNN), tend towards predicting a perpetuation of past observed behavior.
1 code implementation • 25 Dec 2018 • Weijie Jiang, Zachary A. Pardos, Qiang Wei
With cross-disciplinary academic interests increasing and academic advising resources over capacity, the importance of exploring data-assisted methods to support student decision making has never been higher.
no code implementations • 26 Mar 2018 • Zachary A. Pardos, Zihao Fan, Weijie Jiang
The aggregate behaviors of users can collectively encode deep semantic information about the objects with which they interact.
1 code implementation • COLING 2016 • Xingyou Wang, Weijie Jiang, Zhiyong Luo
Sentiment analysis of short texts is challenging because of the limited contextual information they usually contain.