An Automated Fuzzy Active Database for Employee Performance Evaluation Using Oracle
Cyrus Azarbod, Hamed Sallam
Department of Computer & Information Sciences Department Minnesota State University at Mankato, 273 Wissink Hall, Mankato, MN 56001 Cyrus.azarbod@mnsu.edu
Jafar Ali
Quantitative Method & information system department, Kuwait University Kuwait
Abstract
Many corporation and business rely on relational databases for their daily activities and decision making. Employee performance evaluation considered a very time consuming task faced by many large business mangers especially if it involved huge amount of transaction sales. This paper present a methodology based on fuzzy logic techniques to enhance this decision making process. The proposed methodology, involves of building a fuzzy set theory on top of a relational database to automatically generate series of fuzzy tables. These fuzzy tables can be queried for employee performance using SQL (Structured Query Language).
employee performance transactions.
through
sales
2. Methodology
In this paper, we propose a methodology for generating an automated fuzzy employee performance database from an existing sales order relational database. Our methodology takes into account three major components that are considered for integrating fuzzy Logic with relational active database.
2.1. Active Relational Database
Relational database is industry leading database model. It brings simplicity and uses SQL which is a simple and easy to learn querying facility. Active database systems transform passive database processing into active environment by implementing business rules through triggers. A generalized Model that has been used for specifying active database rules is referred to as the Event-Condition-Action, or ECA model.
1. Introduction
Many real business databases deals with a huge amount of data. The large volume and the complexity of databases make it very difficult for management to make a timely decision on important issues such as employee performance. Introducing fuzzy logic technologies to the database fields help in handling uncertain or incomplete information. Combining database and fuzzy technologies create Fuzzy databases. Fuzzy techniques have been applied in many aspects of relational databases such as: representing and querying fuzzy data [1, 3, 6], knowledge discovery from databases [2], and modeling uncertainty at the conceptual schema level [4]. Not a lot of research has been done in generating dynamic fuzzy tables. In this research, we propose a methodology for automatic generation of fuzzy. The proposed methodology will be applied for evaluation
2.2. Fuzzy Logic
Fuzzy logic could assist databases with handling uncertain or incomplete information. Fuzzy set followed by the fuzzification steps used to create the fuzzy tables of the present database. Fuzzy set theory allows the set to have various degrees of membership from 0% to 100% [5]. Fuzzification steps are as follows: 1. Original Table (a1, a2, a3 … an) 2. Let (fi ) be the original table with some fuzzifiable attributes 3. Define the fuzzy functions (categories) for each of (fi) and their fuzzy sets. Choose the suitable values for the fuzzy function parameters
/variables and its range. The Triangular Shape member functions are used to calculate the fuzzy attributes 4. Fuzzy Rules: A single fuzzy if-then rule assumes the form: if x is A then y is B where A and B are linguistic values defined by fuzzy. The input to an if-then rule is the current value for the input variable (in our case, Sales, orders Product,..) And the output is an entire fuzzy set (in this case, Raise, Promotion etc e). This set will later be defuzzified, assigning one value to the output.
triggers and stored procedures will automatically create employee performance table which lead to employee decision table (see Table 1). In step 8, SQL is used to query employee performance and decision tables. For this project, the materialized view refreshes the data every month. Table 1: Employee Performance Decision
EMP_ NO LNAME FNAME REPORT _DATE DECISION
1060 Bixler 1004 Widdes 1069 Alan 1066 Bond 1046 Doering
Charles 11-SEP-05 Fire Albert Dusky Donna 11-SEP-05 Fire 11-SEP-05 Raise 11-SEP-05 Raise Raise and Gift
2.3. Materialized View
Materialized view, introduced with Oracle8i, is designed to improve performance of the database by doing some intensive work in advance before it intended use. Materialized view was used for this project because of its ability to physically create the table (performance consideration) and also its ability to periodically (predefined in advance) refresh the data from new transactions
Thomas 11-SEP-05 Raise Michael 11-SEP-05
1042 Farmer 1047 Constable 1045 Doering … …
Ramond 11-SEP-05 Warning Willard … 11-SEP-05 Warning … …
4. Conclusion
This research successfully has demonstrated on integration of fuzzy logic with active relational database without any human interaction. Powerful relational DBMS such as Oracle can be extended to support fuzzy logic. Seamless extension can provide an exciting new dimension to decision making process that is not available in traditional information systems.
2.4. Steps for building Dynamic Fuzzy Database
Our proposed methodology will create a fuzzy active database through following steps. 1: Determining the fuzzifiable attributes. 2: Defining fuzzy functions (linguistic categories) for each fuzzifiable attributes. 3: Identify the fuzzy range for each category. 4: Identify the membership values for each fuzzy attributes. 5: Build fuzzy tables for each fuzzy attributes. 6: Build performance table to consolidate all fuzzy attributes and fuzzy categories 7: Build decision table for each employee based on the performance table (step 6) 8: SQL query processing using the fuzzy tables.
5. References
[1] Hsieh, N., 2005. Answer to queries concerning uncertain relational databases, International Journal of Intelligent Systems, 20, 647-668. [2] Maddouri, M., Elloumi, S., and Joaua, A., 1998. An incremental learning system for imprecise and uncertain knowledge discovery, Journal of Information Sciences, 109, 149-164. [3] Galindo, J., Aranda, M., Caro, J., Guevara, A., and Aguayo, A., 2003. Applying fuzzy databases and FSQL to the management of rural accommodation, Tourism management, 24, 457-463 [4] Chaudhry, N., Moyne, J., and Rundensteiner, E., 1999. An extended database design methodology for uncertain data management, Information Sciences, 121, 83-112. [5] R.R. Yager et al, 1987 "Fuzzy Sets and Applications: Selected Papers by L.A. Zadeh", ed.. (John Wiley, New York).
3. Implementation
An Order Entry System (OES) active database has been developed using Oracle 9i. OES consists of tables such as Customers, Orders, Product, and Employees. Step 1 through step 4 is done with the management team. Step 5 – 7, is implemented through several materialized views. Any insertion into a fuzzy table will activate triggers to automatically populate fuzzy functions and membership values. Series of