Comparison of decision tree methods in classification of researcher's cognitive styles in academic environment
In today's internet world, providing feedbacks to users based on what they need and their knowledge is essential. Classification is one of the data mining methods used to mine large data. There are several classification techniques used to solve classification problems. In this article, classif...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Penerbit UTM Press,
2015.
|
| Subjects: | |
| Online Access: | Get fulltext |
| Summary: | In today's internet world, providing feedbacks to users based on what they need and their knowledge is essential. Classification is one of the data mining methods used to mine large data. There are several classification techniques used to solve classification problems. In this article, classification techniques are used to classify researchers as "Expert" and "Novice" based on cognitive styles factors in academic settings using several Decision Tree techniques. Decision Tree is the suitable technique to choose for classification in order to categorize researchers as "Expert" and "Novice" because it produces high accuracy. Environment Waikato Knowledge Analysis (WEKA) is an open source tool used for classification. Using WEKA, the Random Forest technique was selected as the best method because it provides accuracy of 92.72728. Based on these studies, most researchers have a better knowledge of their own domain and their problems and show more competencies in their information seeking behavior compared to novice researchers. This is because the "experts" have a clear understanding of their research problems and is more efficient in information searching activities. Classification techniques are implemented as a digital library search engine because it can help researchers to have the best response according to their demand. |
|---|