就業表現與學習歷程的因果規則分析;Complete Causal Rule Analysis for Academic Datasets of Employment Performance and Learning Portfolio

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Computer Science and Information Engineering > Electronic Thesis & Dissertation > Item 987654321/83963

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/83963

Title:	就業表現與學習歷程的因果規則分析;Complete Causal Rule Analysis for Academic Datasets of Employment Performance and Learning Portfolio
Authors:	李睿宸;RUI-CHEN, LI
Contributors:	資訊工程學系
Keywords:	校務研究;因果規則;Institutional Research;Causal Rule Mining
Date:	2020-07-22
Issue Date:	2020-09-02 17:47:07 (UTC+8)
Publisher:	國立中央大學
Abstract:	目前世界各國的許多學校開始重視校務研究，其目的是為了改善教學績效與學校的營運狀，讓決策者透過有效的科學分析，從不同面向分析並做出最適合的對應決策，來面對近年來大專院校不斷增設，衍生出來的許多問題，以更科學化論述事證的方式，來幫助高等教育機構來因應教育環境的變遷與挑戰。校務研究的涵蓋範圍廣泛，包含教務、學務、資源以及人事等，而本研究主要以學生校內行為做為分析面向，根據學生背景的不同，在學時參與的活動、修習、獎懲…等表現亦不同，而這些學生的學習歷程往往會對就業行為有所影響。藉由這些學習歷程資料加上畢業後工作類型與薪資等就業表現，分析學生發展趨勢，對於學校評估學生學習成效和學生活動，或是輔助校務決策有所幫助。本研究以校務資料倉儲作為分析的資料來源，透過循序樣式探勘，探索學生在校狀況及畢業流向的頻繁樣式，去盡可能找尋有可能成為因果規則的序列，並去驗證此規則序列為因果規則的可信程度，以支援校務決策之重要分析。 ;Institutional Research has been a significant feature of higher education in other countries for many years. The main purpose of IR is to improve school operations and teaching performance. Let decision makers have effective scientific methods to analyze and make the most appropriate decision. Help higher education institutions to fight against rapid changes in education policy and society environment. The wide coverage of topic of institutional research includes student affair, academic affair, personnel affair, general affair, etc. This paper focuses on finding causal rules from student portfolios including type of admissions, rewards and punishments, clubs, part-time jobs and questionnaire survey of graduation to analyze employment performance. In order to discover useful causal relations and support decision making, in this paper, data from multi-dimensional data warehouse are seen to employed as input data. Though sequential pattern mining, explore the frequent patterns and obtain association rules. With these association rules, try to find sequences that might be causal rules. Finally, verify the credibility of these rules as causal rules.
Appears in Collections:	[Graduate Institute of Computer Science and Information Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	157	View/Open

社群 sharing

Loading...