Studies and Analysis of Performance Evaluation of Different Student Models with Data Mining Techniques

M. Sridhar, P.V.S. Srinivas, M. Seetha
2017 International Journal of Advanced Research in Computer Science and Software Engineering  
The education environment mostly concerns the outcome based education rather than the physical activity they engage in the learning activity. The outcome of any system is depending on the model of the system. In the present paper various models of EDM are presented and discussed their advantages and disadvantages retrieved from various literatures. A wide variety of EDM models traditional and other advanced models are presented in this paper. Several metrics are available to evaluate the
more » ... ance of different student models. Different metrics are discussed here to support the suitable metric for the student models. In this paper the merits and demerits of mostly used metrics are explored and stated the suitable metric. Keywords --Student models, Metrics, Educational Data Mining, Performance evaluation. I. INTRODUCTION Student performance evaluation is one of the biggest challenges in the present education scenario. Improper student performance evaluation may affect the university ranking, may affect the proposed analysis report prepared for lag students. This report will help the student to take the right decision for their future performance optimization. Based on this report a supervisor can guide the students to enhance their performance in their academic curriculum. The performance evaluation depends on many factors. The factors may influence the performance evaluation process directly or indirectly. Some of inter and intra mutual dynamics influences the performance evaluation. In this context several methods are effectively involved in evaluation process to avoid deviations from the actual requirements. Due to large databases many of the existing techniques failed to analyse the huge student related data. To evaluate large databases an efficient datamining techniques are mandatory to assess a student speculative track perfectly. There are various data mining techniques available to search appropriate record and information from large databases. In the present work various methods are implemented to assess student performance for analysing student achievement and future performance track. Educational Data Mining (EDM) techniques are used to analyse large student data sets. EDM techniques will boost the searching process speed and analysis. In the present work the existing techniques are discussed and presented the defects in various contemporary techniques. The aim of this research is to show the potential of (Educational Data Mining) EDM in enlightening the criteria or measures of effective student performance as perceived by the instructor. The EDM is the branch of soft computing which deals with the concerned data obtained by exploring various education evaluation methodologies to make the students better accustom the current system and to understand the students in the given environment. The EDM is a branch of combination of psychology and learning analytics which are frequently used in analysing student data and reconfiguring the learning model [1] [2] . In recent trends of educational data mining approaches the various algorithms and tools to reach the appropriate student evaluation statistics. The EDM uses various algorithms like machine learning algorithms, data mining techniques and statistical analysis to collect information at various educational and learning schemes. The EDM is a recent research area combination of or mainly concentrated on topics machine learning, data mining, and statistics [3] [4] . New opportunities such as graphics, games, simulations, and tutorials have built in emerging EDM technology to collect student information easily. The data will help the educators to meet their requirements discussed in the above paragraphs. The goals [5] of EDM are classified as predicting students' future learning behaviour, discovering or improving domain models, studying the effects of educational support, and advancing scientific knowledge about learning and learners. For the benefit of reader the merits are presented concisely here. 1. The student performance can be predictable in the given environment. 2. Able to design and modify the student learning methodologies by integrating information such as student knowledge, attitude, motivation etc., obtained from EDM tools 3. Estimation of consequences can be calculated after applying every modification of learning methodologies. 4. It allows reconfiguring advanced Information Communication Technologies (ICT) by building learning and computational model and incorporating knowledge database. Levels of Educational Data Mining As the research area growing day by day variety of data mining techniques came into picture with respect to the context of education environment. After applying each technique the researchers must see that ultimately reach the above goals and benefits. To satisfy the above goals and benefits the EDM is divided into four different levels [1] [6] .
doi:10.23956/ijarcsse/v7i6/0285 fatcat:zx3lsi5o75dmhmrwkuk436xgpu