在資訊化為導向的現代社會中，企業對資料庫的依賴愈來愈深，當資料庫中的某些屬性其屬性值為空值時，將會造成整個資料庫不能正常的運作，此時企業將遭受重大的損失。在本研究中，我們提出一個切割式的機制在人力資源資料庫具有空值時能有效地且更準確地估計空值。我們先採用逐步迴歸來選出資料庫中最重要的兩個屬性，且利用這兩個屬性來建立資料類別。建立好類別之後，針對每一類別中的資料進行分群。接著，計算每一群裡屬性之間的影響程度。本研究中有兩種計算影響程度的方式，一種是使用判定係數（陳鍚明 與 蕭信仁，2004），而另一種則是本研究採用的迴歸係數。利用此兩種方式進而來估計出資料庫之空值。最後，為了驗證我們提出之方法，採用一個實際的資料庫並利用平均絕對誤差率（MAER）來做為評估準則且做為比較之依據。

In general, a database system will not operate properly if it exist some null values of attributes in the system. In this study, we propose a partitional approach for estimating null values in human resource database, which utilize stepwise regression to select the first two important attributes from the database, then partition data by these two important attributes. After building up the data category, we apply the clustering method to cluster data. And next, calculate the degree of influential between the attributes. In this study, there are two ways to calculate the degree of influential. One is using coefficient of determination (Chen & Hsiao, 2004) and the other is using regression coefficients that presented in our study. Furthermore, estimate null values in the human resource database. For verifying our method, this study utilize mean of absolute error rate (MAER) as evaluation criterion to compare with Chen and Hsiao’s method.

1. Introduction 1
1.1 Background and Motivation 1
1.2 Research Objectives 3
1.3 Research Limitations 3
1.4 Organization of the Study 4
2. Literature Review 6
2.1 Review for Estimating Null Value 6
2.2 Stepwise Regression 7
2.3 Correlation Metric 8
2.3.1 Correlation coefficient and coefficient of determination 9
2.3.2 Regression coefficient 10
2.4 Clustering Method 12
2.5 Human Resource Database 14
3. Partitional Approach for Estimating Null Value 17
3.1 Research Framework 17
3.2 Partition Data by Important Attributes 19
3.3 Algorithm Process of Estimating Null Value 21
4. Verification and Comparison 29
5. System Development and Implement 35
5.1 Introduction of the System 35
5.2 The Result of the Experiment 38
6. Conclusions 45
References 47

Arnold, S.F. (1990). Mathematical Statistics. NJ: Prentice-Hall, Englewood Cliffs.
Babad, Y.M. & Hoffer, J.A. (1984). Even no data has a value. Communications of the ACM, Vol. 27, Number 8, August.
Baker, R.E. & Nowak, B. (1997). Statistics on student diversity and multi-culturalism in higher education. Journal of Accounting Education, Vol. 15, No. 1, 71-94.
Bradley, P.S., Fayyad, U., & Reina, C. (1998). Scaling clustering algorithms to large database. The Fourth International Conference on Knowledge Discovery and Data Ming: 27-31.
Chen, S.M. & Yeh, M.S. (1997). Generating fuzzy rules from relational database systems for estimating null values. Cybernetics and Systems: An International Journal 28(2):695-723.
Chen, S.M. & Chen, H.H. (2000). Estimating null values in the distributed relational databases environment. Cybernetics and Systems: An International Journal, 31: 851-871.
Chen, S.M. & Lee, S.W. (2003). A new method to generate fuzzy rules from relational database systems for estimating null values. Cybernetics and Systems: An International Journal 34: 33-57.
Chen, S.M. & Hsiao, H.R. (2004). A new method to estimate null values in relational database systems based on automatic clustering techniques. Information Sciences (In press).
Daniel, W.W. & Terrell, J.C. (1995). Business Statistics for management and economics (7th ed.). Houghton Mifflin Company.
Davidson, J.W. , Savic, D.A. & Walters, G.A. (2003). Symbolic and numerical regression: experiments and applications. Information Sciences 150: 95–117.
Dessler, G. (2002). A Framework For Human Resource Management (2nd ed.). New Jersey: Prentice Hall.
Draper, N.R., Smith, H. (1998). Applied Regression Analysis. New York: John Wiley and Sons.
Dunham, M.H. (2003). Data Mining Introductory and Advanced Topics. New Jersey: Prentice Hall.
Forgy E, (1965). Cluster analysis of multivariate data: Efficiency vs. interpretability of classifications. Biometrics 21: 768.
Ghahramani, S.(1996). Fundamentals of Probability, NJ: Prentice-Hall, Englewood Cliffs.
Hogg, R.V. & Tanis, E.A. (1988). Probability and Statistical Inference (3rd ed.). New York: Macmillan Publishing Company.
Han, J. & Kamber, M. (2001). Data Mining: Concepts and Techniques. Morgan Kaufmann.
Jea, R., Lin, J.L., & Su, C.T. (2005).Correlation and the time interval in multiple regression models. European Journal of Operational Research Volume: 162, Issue: 2, April 16, 433-441.
Lee, J.H., Mao, J.L. (1999). Compensation System and Management (6th ed.). Tsing-Hua Management Science Library.
Laudon, K.C. & Laudon, J.P. (2000). Management information systems :organization and technology in the networked enterprise (6th ed.). London: Prentice Hall.
MacQueen, JB (1967). Some methods for classification and analysis of multivariate observations. Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability 1: 281-297
McClave, T.J., Benson, G.P. & Sincich, T. (2001). Statistics for Business and Economics ( 8th ed.). NJ: Prentice Hall.
Mendenhall, W., Beaver, R.J., & Beaver, B.M. (2001). A brief course in business statistics (2nd ed.). Pacific Grove, CA : Duxbury.
Ng, R.T. & Han. J. (1994). Efficient and effective clustering methods for spatical data mining. Proceddings of the International Very Large Databases Conference: 144-155.
Pitman, J. (1993). Probability. New York: Springer.
Severs, G.L. & Abebe, A. (2004). Rank estimation of regression coefficients using iterated reweighted least squares. Journal of Statistical Computation & Simulation; Vol. 74 , Issue 11, p821.
Zaniolo, C. (1984). Database relations with null values. Reprinted from journal of computer and system sciences 28(1):142-166.