随着一卡通在高校的快速发展，一卡通每日使用次数相当庞大，数据库将一卡通使用次数以数据的形式进行储存。通过对一卡通的使用信息进行分析，可以为校园建设提供指导。虽然已经有研究对这些数据进行分析，但是并没有深入，只是停留在消费数据和消费结构的层面。本次研究通过分析消费与成绩之间的关系，试图找到二者之间的关联性，本次研究的研究课题对于校团委、学生工作处以及教导处均具有非常重要的意义。 本次研究对消费习惯的研究主要以早起和吃早饭的情况出发，针对学生的早餐消费数据和成绩数据，探讨大学生成绩与消费之间的联系。从大学一卡通管理系统ORACLE数据库和教务管理系统MS SQL SERVER数据库中提取样本学生的一卡通消费数据和成绩数据，通过数据筛选和数据清洗等预处理方法对这两种数据进行初步鉴别剔除无效和空白的数据，将通过了预处理的数据以关联算法为主，以K-Means++算法对消费数据进行聚类处理，分析二者之间的关联性，得出了相关结论。总结出大学生消费水平和消费习惯，根据结论对高校对学生的日常学校管理提供指导性意见。基于布尔运算和稀疏矩阵思想，对Apriori算法的缺点进行改进，以提高计算消费数据和成绩数据的效率，从而更高效地发现二者之间的关联性。本次研究所涉及的内容较为新颖，意义重大，为学校管理决策提供科学有效的依据。

With the rapid development of smart card in Colleges and universities, the number of daily use of smart card is quite large. The database stores the number of use of smart card in the form of data. Through the analysis of the use information of one-card, we can provide guidance for campus construction. In the process of using the card, a variety of data will be generated. These data can reflect the situation of college students to a certain extent. Although there have been studies to analyze these data, they have not been in-depth, but remain at the level of consumption data and consumption structure. This study tries to find the correlation between consumption and achievement by analyzing the relationship between consumption and achievement. The research topic of this study is of great significance to the school league committee, student office and teaching office. In this study, the study of consumption habits is mainly based on the situation of getting up early and eating breakfast, aiming at the breakfast consumption data and achievement data of students, to discuss the relationship between college students' performance and consumption. We extract card consumption data and academic management system performance data from ORACLE and MS SQL SERVER databases, and comprehensively process the consumption data and performance data. Based on the huge amount of consumption data, we must preprocess them first and extract the appropriate sample size. The mining technologies of classification, Association and clustering are analyzed and improved, and the performance data and consumption data are analyzed and studied mainly by association algorithm. K-Means++ algorithm is used to cluster the consumption data and summarize the consumption level and consumption habits of College students. Based on Boolean operation and sparse matrix, this paper improves the shortcomings of Apriori algorithm to improve the efficiency of calculating consumption data and performance data, so as to find the correlation between them more effectively. The research involved in this study is relatively new and of great significance, which can better help schools to make scientific decisions and management. This research deeply studies the consumption data of one-card system and the student achievement data provided by the Educational Administration. This research mainly focuses on the study and analysis of students'breakfast consumption habits, explores the relationship between College Students' performance and consumption, summarizes its rules, and provides a scientific and effective basis for school management decision-making.