Educational Data Mining for a Robust Schooling System

**inuwamobarak**

Introduction

Educational data mining (EDM) has become vital for finding knowledge that can be used for different purposes; for example, university management can use knowledge gotten from data of previous students to strategize future plans of the institution. It is a sub-domain of Data Mining that deals with data from academic environments which is used to develop various techniques and to recognize unique patterns.

The knowledge obtained from data can then be utilized by academic planners to provide suggestions that enhance their decision-making. In an effort to improve students’ academic performance, we also decrease failure rates.

In this article, we will see the application of data mining techniques in various academic endeavors. Since many educational activities are being digitized using school portals and databases, a large volume of data containing both students’ data and other data such as staff, lecture time and venues, etc are made available.

A challenge has been to transform the data into knowledge to improve the quality of managerial decisions utilizing techniques such as predicting the academic performance of students at an early stage to help lecturers focus on both excellent students and also identify students with low academic performance.

What is Data Mining?

Data mining can be seen as a field in data science that is concerned with the discovery of novel and potentially useful information from large amounts of data. With data mining, hidden information can be uncovered from large volumes of data. The main goal of data mining is extracting information from a data set and transforming the information into a comprehensible structure for further use.

Types of Data Mining Process

There are different types of data mining processes. We will go through some of them below;

1. Anomaly detection; is also known as outlier analysis. It is a data mining that identifies data points observations that deviate from a dataset’s normal behavior. This involves the identification of unusual data records, that might be interesting or data errors that require further investigation. This could be to search for outliers or a deviation from an expected outcome.
2. Regression; A regression is typically a statistical technique that relates a dependent variable to one or more independent variables. A regression model is able to show a relationship between the dependent variable associated with changes in one or more of the independent variables. It attempts to find a function that models the data with the least error that is, for estimating the relationships in the datasets.
3. Association rule learning; is also known as dependency learning. It involves searching for relationships between variables. For example, a mull might gather data on customer purchasing habits, the supermarket can determine which products are frequently bought together and use this information for marketing purposes.
4. Clustering; is the task of discovering groups and structures in the data that are in some way or another “similar”, without using known structures in the data. It is the task of grouping a set of objects so that objects in the same group (called a cluster) are more similar to each other than in other groups.
5. Classification; The process of assigning components to previously created classes and classifying the resulting set of classes is referred to as classification. It is the task of generalizing known structures to apply to new data. For example, an e-mail program might attempt to classify an e-mail as “legitimate” or as “spam”.
6. Summarization; it provides a more compact representation of the data set, including visualization and report generation.

At the end of all the various steps of knowledge discovery from data, we verify that the patterns produced by the data mining algorithms occur in the dataset. All though not all patterns found by data mining algorithms are in the end valid.

What then is EDM?

From the Educational Data Mining community official website, www.educationaldatamining.org educational data mining is defined as follows: “Educational Data Mining is 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.”

This presents Educational Data Mining as a research field concerned with the application of data mining, machine learning, and statistics to information generated from educational settings.

Some EDM Use Cases

The main purpose of EDM is to analyze and solve educational issues and, consequently, improve educational processes. Several techniques have been identified to implement these applications.

1. Grouping Students; The aim of the grouping students application is to create a group (i.e clustering technique) of students according to different profile information properties or characteristics. This could be grouping students for assignments so that both strong and weak students could complement each other. Some techniques used in grouping students include ANN, and feature selection.
2. Student Modeling; This application uses various variables such as student art skills, emotions, the field of interests, way of learning, past achievements and awards, and learning preferences such as time of reading and way of reading to create a simulation of the environment. Some techniques could be Social Network Analysis (SNA), decision tree, Bayes theorem, and Linear Discriminant Analysis (LDA).
3. Predicting Student Performance; This is one of the most common applications. This is because it tends to be the easiest to apply. The main aim is to predict the student’s academic performance in order to carry out preventive measures early. The most common techniques used are Decision Trees, ANN, Bayesian classification, Rule-based clustering, and feature selection.

EDM Process Life Cycle

The EDM process should follow a set pattern or workflow in other to achieve its goals with more robustness.

Below is a typical EDM Workflow;

Academic Data Collection/Gathering

Data collection is the stage data is acquired for mining. Without this first stage, they can not be EDM. Data may be collected from various sources such as files, questionnaires, school portals, databases, etc. The quality and quantity of gathered data directly affect the result of the desired system.

Academic Data Preparation

Data preparation is done to clean the raw data. Data collected from the university, for instance, is transformed into a clean dataset. Raw data may contain missing values, inconsistent values, duplicate instances, etc. Estimating the missing values for instances using a mean value, median, or mode.

Choosing Learning Algorithm/Technique

This step involves knowing the best-performing learning algorithm to use. It is not best to just choose a model randomly. It depends upon the type of problem that is being solved and the nature of the data set being used. In the case of classification, the data needs to be categorical such as grades (A, B, C), and the data is labeled, and classification algorithms are used.

If the problem is to perform a regression task with a continuous output like marks and the data is also labeled, regression algorithms will be suitable. If the problem is to create clusters and the data is unlabeled, clustering algorithms such as K Means can be used.

Training/Testing Model

The model training phase involves training the model to learn from the acquired clean data with the selected algorithm. For a classification or regression problem, the student dataset is divided into a training and a testing dataset.

The training dataset is used for training purposes. The testing dataset is used for testing purposes, to know how currently the model has learned, while the training dataset is fed to the learning algorithm.

Evaluating the Model

After we have gotten a model result, from the trained and tested algorithm, the evaluation of the model validates the model for effectiveness. The model is evaluated using the test dataset and is seen to be sure it meets production requirements.

Metrics such as precision, prediction, recall, etc are used to test the performance. During prediction, the built model is used to do something useful in the real world.

Conclusion

In this article, we have seen the application of data mining techniques in various academic endeavors. Since many educational activities are digitized, such as portals and databases, a large data is available making it possible for the Educational Data Mining field to strive.

One of the most common activities in EDM has been to predict student academic performance. Because this helps improve the performance of institutions. With the current trend in online educational systems, EDM is a perfect field to specialize in.

Key Takeaways:

• Data mining can be seen as a field in data science that is concerned with the discovery of novel and potentially useful information from large amounts of data.
• Educational Data Mining as a research field is concerned with the application of data mining, machine learning, and statistics to information generated from educational settings.
• The main purpose of EDM is to analyze and solve educational issues and, consequently, improve educational processes.

Links;

www.educationaldatamining.org

https://www.freepik.com/free-photo/shallow-focus-shot-african-child-learning-school_14376285

https://www.freepik.com/free-vector/geometric-science-education-background-vector-gradient-blue-digital-remix_17213260

https://www.pngwing.com/en/free-png-bqqqq/download

https://www.pngwing.com/en/free-png-mrama/download