隨著高等教育日益重視學用合一，學生在學期間的表現是否會影響其畢業後進入職場的職涯發展，已成為教育政策與學生職涯輔導領域的重要議題。然而，現有研究多著重於挖掘正向關聯規則，較少關注主修學系與職涯發展方向不一致的負向關聯規則。為了更了解學生畢業後的職涯發展選擇，本研究透過挖掘正負向關聯規則，並結合勝算比進行關聯規則強度分析，探討學生在學表現與畢業後職涯發展方向的潛在關係。
本研究資料來自國立中央大學校務研究資料倉儲，資料集包含畢業生基本資料及畢業後滿一、三、五年的職涯發展問卷資料。使用的在學表現屬性為畢業學院、畢業學系與畢業成績，職涯發展屬性為職業類型、學系與工作相符程度、學系與工作的相符程度、具備能力與工作要求相符程度等。本研究應用FP-Growth與E-NFIS（Efficient Negative Frequent Itemsets Mining）進行正負向關聯規則探勘，並限制條件（Antecedent）為在學表現屬性，結果（Consequent）為職涯發展屬性，以確保時間序與推論合理性。
研究結果顯示，正向關聯規則如資訊電機學院學生較常投入科技相關產業，反映學生主修與職業之間具一定對應性；同時，負向關聯規則補充了正向關聯規則未能涵蓋的面向，如部分學系學生較少進入與主修相關領域工作。同時挖掘正負向關聯規則有助於減少片面解讀的可能，並可為課程設計與職涯輔導策略提供參考依據。

With the growing emphasis on aligning academic learning with practical application in higher education, the extent to which students' academic performance influences their post-graduation career development has become a key issue in education policy and student career counseling. However, most existing studies have primarily focused on mining positive association rules, with relatively little attention paid to negative association rules that reveal discrepancies between students’ academic majors and their career development paths. To gain a more comprehensive understanding of graduates' career development, this study investigates the potential relationships between academic performance and career trajectories by mining both positive and negative association rules and analyzing the strength of these association rules using odds ratios.
The data for this study were obtained from the Institutional Research Data Warehouse of National Central University. The dataset includes graduates’ basic profile data as well as career development survey responses collected one, three, and five years after graduation. The academic performance attributes used in this study include college of graduation, department, and final academic performance, while the career development attributes include job type, the degree of relevance between the job and the major, the consistency between the job and the academic department, and the match between the required competencies and job requirements. The study employed FP-Growth and E-NFIS (Efficient Negative Frequent Itemsets Mining) algorithms to extract both positive and negative association rules, with antecedents restricted to academic performance features and consequents to career development features to ensure temporal sequence and logical reasoning.
The research findings indicate that positive association rules—such as the tendency of students from the College of Electrical Engineering and Computer Science to enter technology-related industries—suggest a certain degree of alignment between students' academic majors and their career paths. At the same time, negative association rules complement aspects not captured by positive rules, such as the observation that graduates from certain departments are less likely to work in fields directly related to their majors. The simultaneous mining of both positive and negative association rules helps reduce the risk of one-sided interpretations and can inform curriculum design and career counseling strategies.

[1] M. Salas-Velasco, "Mapping the (mis)match of university degrees in the graduate labor market," Journal for Labour Market Research, vol. 55, art. 14, 2021.
[2] R. A. Voorhees, "Institutional Research's Role in Strategic Planning," New Directions for Higher Education, no. 141, pp. 77-85, Spring 2008.
[3] L. Zhang, "Association Rule Mining for Career Choices Among Fresh Graduates," Applied and Computational Mathematics, vol. 8, no. 2, pp. 37, 2019
[4] T. Wang, B. Xiao, and W. Ma, "Student behavior data analysis based on association rule mining, " International Journal of Computational Intelligence Systems, vol. 15, p. 32, 2022.
[5] I. Papadogiannis, M. Wallace, and G. Karountzou, "Educational data mining: A foundational overview, " Encyclopedia, vol. 4, no. 4, pp. 1644–1664, 2024.
[6] R. Agrawal, T. Imieliński, and A. Swami, "Mining association rules between sets of items in large databases," ACM SIGMOD Record, vol. 22, no. 2, pp. 207-216, 1993.
[7] P.-N. Tan, V. Kumar, and J. Srivastava, "Selecting the right objective measure for association analysis," Information Systems, vol. 29, no. 4, pp. 293-313, 2004.
[8] J. L. Saupe, The Functions of Institutional Research, 2nd ed. Association for Institutional Research, 1990.
[9] J. F. Volkwein, "The foundations and evolution of institutional research," New Directions for Higher Education, vol. 2008, no. 141, pp. 5-20, 2008.
[10] K. L. Webber, "Institutional research and decision support in higher education: Considerations for today and for tomorrow," in Building Capacity in Institutional Research and Decision Support in Higher Education, K. Webber, Ed. Springer, 2018, pp. 1-14.
[11] W. H. Inmon, Building the Data Warehouse. John Wiley & Sons, 2005.
[12] P. J. Goldstein and R. N. Katz, Academic Analytics: The Uses of Management Information and Technology in Higher Education. EDUCAUSE, 2005.
[13] G. Siemens and R. S. J. d. Baker, "Learning analytics and educational data mining: Towards communication and collaboration," British Journal of Educational Technology, vol. 43, no. 3, pp. 389-396, 2012.
[14] B. Daniel, "Big data and analytics in higher education: Opportunities and challenges," British Journal of Educational Technology, vol. 46, no. 5, pp. 904-920, 2015.
[15] J. Han, M. Kamber, and J. Pei, Data Mining: Concepts and Techniques, 3rd ed. Morgan Kaufmann, 2011.
[16] R. E. Walpole, R. H. Myers, S. L. Myers, and K. Ye, Probability and Statistics for Engineers and Scientists, 9th ed. Pearson, 2012.
[17] R. Agrawal and R. Srikant, "Fast algorithms for mining association rules in large databases," in Proc. 20th Int. Conf. Very Large Data Bases (VLDB), 1994, pp. 487-499.
[18] J. Han, J. Pei, and Y. Yin, "Mining frequent patterns without candidate generation," in Proc. 2000 ACM SIGMOD Int. Conf. Management of Data, 2000, pp. 1-12.
[19] M. Al-Maolegi and B. Arkok, "An improved Apriori algorithm for association rules," International Journal on Natural Language Computing (IJNLC), vol. 3, no. 1, pp. 21-29, 2014.
[20] X. Wu, C. Zhang, and S. Zhang, "Efficient mining of both positive and negative association rules," ACM Transactions on Information Systems (TOIS), vol. 22, no. 3, pp. 381-405, 2004.
[21] M. L. Antonie and O. R. Zaïane, "Mining positive and negative association rules: An approach for confined rules," in Knowledge Discovery in Databases: PKDD 2004, J. F. Boulicaut, F. Esposito, F. Giannotti, and D. Pedreschi, Eds. Springer, 2004, pp. 27-38.
[22] X. Dong, L. Ma, and X. Han, "E-NFIS: Efficient negative frequent itemsets mining only based on positive ones," in Proc. IEEE 3rd Int. Conf. Communication Software and Networks, 2011, pp. 517-519.
[23] F. Hao, L. Zhao, T. Xu, and X. Dong, "Application of negative and positive association rules in mental health analysis of college students," in Proc. 14th Int. Conf. Natural Computation, Fuzzy Systems and Knowledge Discovery (ICNC-FSKD), 2018, pp. 1462-1467.
[24] J. M. Bland and D. G. Altman, "The odds ratio," BMJ, vol. 320, p. 1468, 2000.
[25] A. Agresti, Statistical Methods for the Social Sciences, 5th ed. Pearson, 2018.