Graduate Enrollment Management System Design And Outlier Detection Implementation

With the increasing expansion of graduate education in China, the roll management of graduate students is faced with many new problems, calling for a scientific and information-based management. Currently, the graduate student roll management in our university primarily adopts two ways, namely manual management mode and decentralized, independent system. Different departments use stand-alone management software independently, most of which have unitary function and unfriendly operating interface. Moreover, the complex maintenance of the software and the existence of data islands make it difficult to integrate and synchronize data completely. There is no interface among different software, resulting in plenty of repetition in data entry and more workload of the Graduate School. This way of working is not efficient and we can not take full advantage of the student roll data. So it cannot meet the growing demand. In addition, the current graduate student roll data is not effective enough, and there are problems in terms of student roll registration and attendance, which have affected the quality of graduate student management to a certain extent. Therefore, this study intends to establish a new, unified graduate student roll management system, to adopt LBS technology and abnormal data detection algorithm to solve the problem of difficult access to data in postgraduate student roll management, to accomplish automation and modernization of student roll management, and to provide greater technical support for management decisions.This paper is based on the management practice of the Graduate School of Northeastern University. By analyzing its management process and using advanced design concepts, a graduate student roll management information system covering various aspects of graduate management has been developed, so as to achieve the integration and data sharing among subsystems, including enrollment, student roll, training, performance, curriculum, lecturers, graduation and so on. The system has realized abnormal data detection in all aspects of graduate student management, from admission to graduation. It has satisfied the demand of users in graduate student information management, and has effectively solved a series of practical problems, including data repetition, waste of resources, difficult access to student roll management data, extremely large amount of data, and so on. In this way, it has achieved an information-based, transparent management process in education and teaching. This paper presents an abnormal detection method for graduate student roll data. This method can effectively and actively detect the change in student roll, performance distribution, and attendance abnormality. The system design in this paper adopts LBS-based electronic attendance technique and abnormal data detection technique, effectively reducing the workload of staff and improving efficiency. In the process of graduate student roll management, it will proactively identify the abnormal data in student roll management, so that the managerial staff can achieve effective management and automated management of graduate students in the entire process. This plays an important role in improving the overall quality of graduate student training and the level of graduate education in our university.