Data Mining for Learning and Study Strategies Inventory

• Main Authors: Yan-Zhen Lee, 李延震
• Other Authors: 蔣定安
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/93567384064106075204

碩士 === 淡江大學 === 資訊工程學系碩士班 === 94 === When we wish to understand the thought processes and psychology of a large number of people, we often utilize surveys to obtain objective data. These surveys, if they are structured in a thorough manner, are frequently composed of dozens of topic questions that can be quite time-consuming for the subject to complete and may therefore be looked upon by the subject with a feeling of dread. To avoid this sort of a situation, we have come up with a method of extracting valuable answers from surveys using the classification charts of decision tree analysis to evaluate the subject through a lesser number of more relevant questions. Once certain indicators have been pinpointed, associated rules can then be used to assess other indicators by deciding whether they can be assessed by the original indicator. We can then utilize our assessments of several indicators in conjunction with one another to decide whether another level of assessment is needed. It is possible in this way to reduce the number of topic questions while obtaining approximate results, and the simplified topic questions do not then need to be posed through a formal survey but can rather be presented in a regular dialogue or interview. Preliminary data can then be obtained without causing the subject to shrink away from the task and subjects that experience difficulties studying and stress at handling their problems can be pinpointed. It is possible then to apply the analysis results to the formula and provide the subject with an outlet for self-assessment. In addition, once the results of this self-assessment have undergone statistical analysis, the data for users that need to seek counseling can be provided to school counselors who can then provide yet another level of counseling. In this paper, we first use IM8.1 (Intelligent Miner for Data 8.1) as a model for this data (a questionnaire concerning learning and study strategies) to draw up the analytical operations for decision trees and association rules. Those rules can then be inputted into the database. Most users of the ASP design user interface can perform self-assessment using this system, and school counselors can inspect counseling results by looking at the equations and acquire survey results.