Data Mining Methodology

1429 Words6 Pages
Educational Data Mining (EDM) is a developing field based on data mining techniques. EDM emerged as a combination of areas such as machine learning, statistics, computer science, education, cognitive science, and psychometry. EDM focuses on learner characteristics, behaviors, academic achievements, process of learning, domain knowledge content, assessments, educational functionalities, and applications. Educational data mining is defined by Baker (2010) as ‘‘an emerging discipline, concerned with developing methods for exploring the unique types of data that come from educational settings, and using those methods to better understand students, and the settings which they learn in’’. EDM is concerned with improving the learning process and environment.…show more content…
With Student Modelling, learning environments can be designed according to learners’ needs and expectations. This part consists five chapters which are focused on developing systems for identifying learners and estimate their performance automatically with machine learning especially in intelligent tutoring systems and adaptive learning environments. Learner attributes such as learner characteristics, behaviors, learning styles, educational experiences, personality, academic achievements, learning system usage data, and assessment results are used for the development of these systems. It is aimed to assist tutors and to increase quality of education, learners’ satisfaction and achievement with these systems developed for performance estimation. It is emphasized that estimating performance is especially important for determining at-risk students in the first years and for retention of those students. It is stated that student models can be used in all learning environments. However, using them especially in e-learning systems in which learners and instructors are not at the same place can be more effective. In the fourth Chapter, knowledge discovery processes and development of Student Knowledge Discovery Software which is created with determination of learner features has been explained. Student performance is modeled using data mining techniques. In the fifth…show more content…
In the assessment process, learners’ domain knowledge acquisition, skills development, and achieved outcomes are taken into account. In addition to these, it is underlined that reflection, inquiring, and sentiments are significant for computer based educational systems. In the ninth chapter, a coherence analyzer is designed to be used in an Intelligent Tutoring System. With this kind of work, learners evaluate their drafts early, and facilitate the reviewing process of the academic advisor. The tenth chapter offers an approach to create test automatically. In this chapter, a model has been developed by using EDM techniques to estimate the difficulty of the items in the computer-based tests. In the eleventh chapter, an instrument has been developed in Exploratory Learning Environments to support teachers’ understanding of students’ activities and visualize these activities. In the study, novel methods and algorithms for augmenting existing pedagogical software for science education has been developed. In the twelfth chapter, a new approach to find the most dependent test items in students’ response data by adopting the concept of entropy from information theory has been proposed. It is highlighted that students’ response data can be used to determine the knowledge learned by students and it can also be used to discover the relationship between the test items
Open Document