The ultimate goal of English teaching is to cultivate the students’ ability to communicate information in English, master good language learning methods, and become independent language learners and users. Therefore, successful English language teaching needs to be achieved through language communication training between teachers and students and between students. This article investigates the importance of promoting the reform of oral English teaching in China’s English teaching environment. We believe that to promote the reform of oral English teaching, an oral teaching environment must be available. However, the current common problem in oral English teaching in colleges and universities is that the spoken conversation objects are not standard enough, or there is no person who can talk to. Therefore, an intelligent spoken dialogue system based on big data and neural network technology is particularly important, and the quality of dialogue depends on accurate spoken speech evaluation. We first extracted six features of pronunciation quality, fluency, content richness, topic relevance, grammar, and vocabulary richness. Secondly, we propose an evaluation model that connects specific TDNN layers in a feedforward manner, using the feature representation of target words in different TDNN layers, which can obtain richer context information and greatly reduce the amount of model parameters. Finally, we conducted a simulation experiment. The experimental results show that the proposed model is accurate in evaluating spoken English and can effectively assist the reform of spoken English teaching in colleges and universities, and its performance is better than SVM by 9.2%.

中文翻译：

---

大数据和深度神经网络技术驱动的大学英语口语教学改革

英语教学的最终目标是培养学生用英语交流信息的能力，掌握良好的语言学习方法，成为独立的语言学习者和使用者。因此，成功的英语教学需要通过师生之间、学生之间的语言交流训练来实现。本文探讨了在我国英语教学环境中推进英语口语教学改革的重要性。我们认为，要推进英语口语教学改革，必须要有口语教学环境。然而，目前高校英语口语教学普遍存在的问题是口语会话对象不够规范，或者没有可以交谈的人。所以，基于大数据和神经网络技术的智能口语对话系统尤为重要，对话的质量取决于准确的口语评价。我们首先提取了发音质量、流畅度、内容丰富度、主题相关性、语法和词汇丰富度六个特征。其次，我们提出了一种以前馈方式连接特定TDNN层的评估模型，利用不同TDNN层中目标词的特征表示，可以获得更丰富的上下文信息并大大减少模型参数量。最后，我们进行了模拟实验。实验结果表明，该模型对英语口语评价准确，能有效辅助高校英语口语教学改革，