Recommendation Algorithms for Improving University Lectures

摘要

The present invention relates to a recommendation system for improving university lectures, so that students who have taken the course based on the everyday lecture evaluation structured data provide evaluations on total points, assignments, group meetings, grade point ratio, attendance, number of tests, and semesters taken, It allows students to provide subjective general reviews based on time course evaluation unstructured data, analyzes and predicts grade evaluation patterns based on in-school undergraduate student data, and collects and analyzes unstructured data on lecture evaluation by a search engine. The lecture method recommendation algorithm is used to determine which part should be improved for the professor with the low total score by using the lecture overall rating of the subject with the high total score among the structured data variables of the everyday lecture evaluation by the lecture improvement recommendation system. Through the data on lecture reviews, students decide whether to take the lecture or not, but the lectures of professors through this lecture review are not improved. However, it is intended to solve the problem that professors' lectures through this lecture review do not improve. That is, the present invention is everytime lecture evaluation formatted data configured in a certain format so that students who have taken in the university community system (everytime) provide evaluations on total points, assignments, group meetings, grade point ratio, attendance, number of tests, and semesters taken. , Collecting unstructured data for everytime lecture evaluation in an irregular format so that students who have taken the course can provide a subjective overall evaluation, intra-school undergraduate student data in a uniform format to analyze and predict grade evaluation patterns, and unstructured data on lecture evaluation and a search engine composed of Python crawling code for analysis, everytime lecture evaluation structured data variables, using the lecture summary of subjects with high total scores in the total score section to recommend lecture methods to professors with low total scores on which areas to improve It is composed of a lecture improvement recommendation system formed in the form of an app to provide an algorithm. Therefore, the present invention allows students who have taken classes based on the Everytime Lecture Evaluation structured data to provide evaluations on total points, assignments, group meetings, grade point ratio, attendance, number of tests, and semesters taken, and take classes based on the Everytime Lecture Evaluation atypical data. Allow students to provide a subjective overall evaluation, analyze and predict grade evaluation patterns based on intra-school undergraduate student data, collect and analyze unstructured data on lecture evaluation by a search engine, and use a lecture improvement recommendation system By using the overall lecture evaluation of subjects with a high total score among all-time lecture evaluation structured data variables, the lecture method recommendation algorithm is provided to the professor with a low total score on which part to improve. Through data, students decide whether to take a lecture or not, but by solving the problem that professors' lectures do not improve through this lecture review, students decide whether to take the lecture through the data on the existing lecture review, but this lecture review The lectures of professors through the system have the effect of solving problems that cannot be improved.