Frequent Pattern Mining for Educational Data By Using Mapreduce Approach In Hadoop

Abstract

Lots of structured and unstructured information can provide educational insights through information retrieval. Gathering educational information in an effective way is also a major challenge. In addition, historical records can be obtained by effectively storing the information collected. Traditional analytical methods need to be expanded. The proposed method can effectively manage large volumes of data by combining apriori, prefix tree and eclat solutions. These techniques are reported to be more effective by integrating Hadoop and Mapreduce platforms. It also builds educational data collection software on Android mobile phones designed to improve education in the home country. This software can be accessed by adding a user account. A QR code is used to verify that you are a registered student or teacher. In the education area of Myanmar, computers, mobile and internet have become important tools for high school students. To enable the quality and the flexibility of the education, verities of education programs and methods are greatly included but with different manners. Frequent itemset mining (FIM) is most popular technique in datamining area like health care, manufacture and finance. The proposed system develops two parts. The first part is implements teacher assessment survey application. Teacher assessment survey application is to collect educational data. The second part is introduced two FIM methods. These are Apriori Prefix Tree Eclat (ATE) and Eclat Prefix Tree (ET) based on Hadoop Mapreduce platform. Proposed method ATE is developed for large dataset to handle scalability and optimization. ET method is implemented to compare compile time to proposed method ATE. The proposed method is thinking instructor, student behavior and giving managerial decision support, analysis on teacher assessment survey data. The proposed method is for the decision maker. It helps to think about supporting student behavior and management decisions. It can be used to manage the frequent itemset results of the proposed method. It can effectively analyze large pieces of educational data in a timely manner. Using this proposed method can be used to effectively analyze the frequency and results of educational data in a short period of time. In addition, the proposed method is applicable to many universities. For university can also be used in management for individual teachers or for multiple teachers. It is particularly suitable for the analysis of large educational data.