随着计算机技术的飞速发展与广泛应用，教育领域面临着如何有效利用收集到的学生信息数据集这一课题，其中一个重要的研究方向是如何利用机器学习算法预测学生语言表达能力。本文将基于不同年级、不同性别的学生口述文本的数据，对文本进行分句分词、词性标注等预处理，提取文本的特征向量，以人工标注的表达能力为标签，训练支持向量机、决策树分类、逻辑回归和多层感知机这四种机器学习模型。通过对比不同机器学习模型的评估指标，筛选出在学生语言表达能力评分的方向下综合水平最优秀的机器学习算法。本研究训练出的机器学习模型不仅可以提高评分的效率和一致性，还能帮助教师更高效地进行教学评估，同时也能够为学生提供个性化的学习建议和学习指导。 

With the rapid development and wide application of computer technology, the field of education is faced with the problem of how to effectively use the collected student information datasets, and one of the important research directions is how to use machine learning algorithms to predict students' language expression ability. Based on the data of students' oral texts of different grades and genders, this paper will preprocess the text with sentence segmentation and part-of-speech annotation, extract the feature vectors of the text, and train four machine learning models, namely support vector machine, decision tree classification, logistic regression and multilayer perceptron, with the expression ability of manual annotation as the label. By comparing the evaluation indicators of different machine learning models, the machine learning algorithm with the best comprehensive level in the direction of students' language expression ability score was selected. The machine learning model trained in this study can not only improve the efficiency and consistency of grading, but also help teachers conduct teaching assessments more efficiently, and also provide personalized learning suggestions and learning guidance for students.