Discussion

1. What is the definition of data mining that the author mentions?

According to the article, data mining involves recognizing data trends and predicting models for analyzing data using data analysis methods to find unseen data information (Križani?, 2020). He also says that data mining is broadly known as studying vast data to get the knowledge one needs. In addition, data mining focuses on finding parameter relationships for massive data types. The author also presents data mining as getting relevant data information and know-how from a data warehouse. Moreover, he says that it involves employing various tools to analyze data to sense unknown relationships and data patterns. Finally, he compares data mining to methods of inventing an automatic process; it involves extracting relevant information from raw data s that such data is absorbed in process mining.

2. How is this different from our current understanding of data mining?

Currently, experts define data mining as a process of discovering knowledge in data. It involves finding patterns and necessary information from massive data sets to aid in developing machine learning algorithms (Delen, 2021). This definition is similar to the author's, but the difference comes in the data mining purpose.  With technological advancement, most organizations do data mining to use machine learning technology to predict future trends in the current data. The author’s definition focuses only on finding relationships and patterns in data. However, later in the article, he mentions machine learning; thus, it is may not be appropriate to say that the two definitions are far distinct.

3. What is the premise of the use case and findings?

The methodology and results of the research depend on decision tree and cluster techniques that the researcher used in educational data mining. The author used these strategies to derive conclusions based on research from a Croatian higher education institution from February to June 2018. The study involved getting information on students’ behavior towards the e-learning system and how the traits affected their results. Also, based on the application of RapidMiner in data analysis, the author was able to get a visual representation of findings as decision trees and clusters for result validation. It is similar to simulating a process to predict its transition before testing it (Križani?, 2020).

4. What tools are in the data mining aspect of the use case, and how are they used?

The data mining tool in the article is the decision tree and cluster analysis software, RapidMiner. The research employs the decision tree software to analyze results from clusters. In this case, a single decision tree represented analysis from each cluster group. They used cluster analysis to group students with identical behaviors towards the electronic system together. Moreover, after clustering, the research found three groups of students for this case. The three clusters meant that there were also three decision trees (Križani?, 2020).

The RapidMiner tool employed the K-means algorithm for cluster analysis and Z-transforms for normalization. The clustering variables were students’ identification documents. Types of measures for group formation were numerical values of Euclidean distance. The number of times each student accessed the materials were the influential variables. On the other hand, RapidMiner’s settings in making decision trees focused on predicting students' number of points in the exams. Here, the students’ points were the high gain varying feature. The normalization technique was the same as that of clustering. The strategy for tree creation was the least square (Križani?, 2020).

 

5. Were the tools used appropriately for the use case? Why or why not?

RapidMniner studio is software for the analysis of processes through graphical designs and other visual representations. This tool can also handle any data type: whether structured or unstructured (Tahyudin, 2015). This case includes different kinds of data such as frequencies of access to the e-content, students' IDs, and student's points in exams. Since RapidMiner can handle a variety of data, it was significant in this research's analysis. The frequencies and examination points were in numerical forms hence easy to quantify using the software. Also, instead of using student's names, the researcher used student's identification documents. The IDs were numbers; hence were easy for analysis (Križani?, 2020). In addition, through decision trees development and grouping, we can see a visual representation of the data. Finally, by clustering and decision trees, we can find types of groups available in the institution based on the students' characters.

The use of the above tool gave convincing results. However, using the frequency of a student’s access to the e-content to predict the students’ performances in the future may not be reliable. This unreliability is a result of behavioral change among human beings. For instance, a future student may access the e-learning portal frequently but fail in exams. Another student may access the online site a few times but pass the exams. In such instances, the results of an analysis may be wrong, and using this tool may not be appropriate. It is, therefore, valid to say that such a tool maybe not be applicable to predict the behavioral impact on students' results in the future based on current findings.

 

Title: Data Mining_ Krizanic (2020)

Thesis statement: Authors define data mining differently. Though most of the definitions differ, they make similar sense to any reader. The distinction is only in the wording. Also, in the case of studies, researchers use different approaches and tools to mine data from various sources. The methodologies and tools that they use may be appropriate in some cases but not in others.  

1. Introduction

1. Authors define data mining differently. Though most of the definitions differ, they make similar sense to any reader. The distinction is only in the wording. Also, in the case of studies, researchers use different approaches and tools to mine data from various sources. The methodologies and tools that they use may be appropriate in some cases but not in others.

 

2. Author’s definition of data mining
3. Author's definition's difference from current understanding
4. The premise of the use case and findings
5. Type of tools used in the use case and how they were are they used
6. Were the tools used appropriately for the use case? Why or why not?
7. Conclusion

1. Authors define data mining differently. Though most of the definitions differ, they make similar sense to any reader. The distinction is only in the wording. Also, in the case of studies, researchers use different approaches and tools to mine data from various sources. The methodologies and tools that they use may be appropriate in some cases but not in others.