Shuffling Algorithms for Automatic Generator Question Paper System

Examination process is important activities for educational institutions to evaluate student performance. Thus the quality of the exam questions would determine the quality of the students produced by the institutions. Preparing exam questions is challenges, tedious and time consuming for the instructors. Usually the instructors keeping their own test bank in some form to help them prepare future exams. Current technologies help the instructors to store the questions in computer databases. The issue arise is how the current technologies would also help the instructors to automatically generate the different sets of questions from time to time without concern about repetition and duplication from the pass exam while the exam bank growing. This paper describes the usage of shuffling algorithm in an Automatic Generator Question paper System (GQS) as a randomization technique for organising sets of exam paper. The results indicate shuffling algorithm could be used to overcome randomization issue for GQS.


Introduction
Producing a quality graduates is one of the main objective in any educational institution.The higher acceptance of their graduates in work market indicates the quality of the institutions.The quality of graduates produced by any institutions is determined by many factors.One of the factors is the quality of the evaluation system.Evaluation system could exist in many forms.Conventional evaluation system is normally based on the exam system.Before the exam could be given to the student, the instructors or lecturers must prepare the questions according to the topics covered for each of the subject.Preparing exam questions is challenges, tedious and time consuming for the instructors.Usually the instructors keeping their own test bank in some form to help them prepare future exams.Current technologies help the instructors to store the questions in computer databases.The issue arise is how the current technologies would also help the instructors automatically generate the different sets of questions from time to time without concern about repetition and duplication from the pass exam while the exam bank growing.

Current System of Generator Question Paper
There are a few systems in today market that offered the similar services like what describes in this paper.These systems are developed by different developer with different features.The randomization techniques employed by such system is also different.Below is one of current system.

ASTech's JARPPL System
Atlantic Simulation and Training Tecnologies (ASTECH, 2010) introduced automated examination management system known as JARPPL.JARPPL features as describes in ASTECH includes the entire basic requirement essential for managing examination process.The main function includes importing of the generating MCQ (Multiple Choice Questions) exams, automatic exam correction, storing results and producing reports.
ASTech's JARPPL System provides an intuitive interface and incorporates a main menu allowing access to any of the functionality at any stage.An On-Line help file, accessible from the main menu, ensures that instructions on using the system are always accessible.
The main functionality of the system as summarised in their web page is as follows: i.
Generate draft examinations by randomly or manually selecting questions from the central question databank.ii.Sign-off and prepress.iii.Print individualised exam papers and solution sheets iv.Maintain complete records of applicants (exam results, dates, etc.) v. Produce school, NAA and JAA pro forma reports.

Generator Question Paper System (GQS)
This paper introduced the usage of shuffling algorithm in Automatic Generator Question Paper System (GQS) to overcome the issue stated.The main role of the shuffling algorithms is to provide randomization technique in GQS thus different sets of question could be generated without repetition and duplication.Randomization technique is a method that has grown enormously over the past 20 years (Ian, 1998).This technique in widely use in gambling field.There have four categories of randomization which are generating a random permutation of a sequence, selecting a random sample of a population, generating random numbers and transforming a data stream using a scrambler in telecommunications (Wikipedia, 2010).
GQS is a special online system which generates question paper or test papers of user's choice just on button click based.GQS is the system that has features for keeping test bank question and produced exam paper.This is automatic process of exam paper generation together with answering scheme.It uses a huge question bank with based on the learning outcomes elements which refer to the Bloom's Taxonomy.Bloom's Taxanomy includes 6 elements of learning outcomes which are knowledge, comprehension, application, analysis, synthesis and evaluation.Bloom's Taxonomy (Barbara, 1994) divides the way people learn into three domains.One of these is the cognitive domain which emphasizes intellectual outcomes.This domain is further divided into categories or levels.The key words used and the type of questions asked may aid in the establishment and encouragement of critical thinking, especially in the higher levels.This system is very useful for lecturers.The system is available for users to generate the questions of all the subjects of exams.Lecturers also will free from tension of collecting question to generate question paper for examination.
The functions in GQS are embedded with learning outcomes measure that would help lecturers produced quality exam question according to learning outcome objective for each course.This system is being tested for Faculty Computer Science and Information Technology (FSKTM) in Universiti Putra Malaysia.The questions are stored in huge Question bank.Lecturers can add new questions, delete the old questions and update the existing question in the Question bank at anytime.Lecturers could also generate different sets of question papers from the same database with just one click by selecting all the requirements needed.This system will prevent duplication of the questions by using the shuffling algorithms.The option to choose shuffling algorithm for randomization is because of simplicity.Apart of shuffling algorithm features are to prevent duplication and repetition of generated sets of question paper.All the questions are picks randomly from the database.Lecturers also can save generated papers for later use or delete after print.

Shuffling Algorithms Implementation.
Shuffling algorithms is very suitable and effective ways to implement for randomization of generated question.This algorithms check for duplication and repetition of the randomize question.The behaviour of the algorithm is as followed, for a set of N (the total number of question in the database) elements for generating a random permutation of the numbers 1-N goes as follows (wikipedia, 2010c) 1) Select the numbers from one to N in the database.
2) Pick a random number k between one and the number of unstruck numbers remaining (inclusive).
3) Counting from the low end, strike out the kth number not yet struck out, and mark it.4) Repeat from step 2 until all the numbers have been struck out.
5) The sequence of numbers written down in step 3 is now a random permutation of the original numbers.

Findings and Discussions
The main objectives of the system are: 1) To applied randomization technique for the Automatic Generator Question Paper System by using shuffling algorithms.
2) To help lecturers preparing set of question paper based on the learning outcomes elements.

Randomization Achievement
The first objective has been achieved when generating 5 sets of question paper with same attributes.Result shown the questions are picking up randomly in the database according to the sequences.All 5 sets question papers generated by same lecturer using GQS.The lecturer generated 5 sets with same attributes and characteristics with 30 questions each set.The shuffling algorithms demonstrated that the entire question in 5 different sets is generated randomly.The result also indicates no duplication and repetition of the produced question.The shuffling algorithm employed shows it effectively support the randomization techniques especially in selecting the random data in one population.
In the implementation of GQS, The system successful generated random questions paper.Table 1 show that the first objective is achieved.(Note 1)

Randomization Based on Learning Outcomes Elements
GQS is an automatic generator question paper which is based on the learning outcomes elements.All the questions prepared are accompanied by their set of learning outcomes to be achieved by each of the respective topics.
The second objective for GSQ which is to help lecturers preparing set of question paper based on the learning outcomes elements.This objective has been achieved when generated 3 types of different question paper which are test 1, test 2 and final examination.All the question paper has 20 questions per set.The question paper that had been generated shown that all the questions generated are followed the rules that randomly selected based on the learning outcomes elements.Figure 2 shows all the questions generated are randomly selected based on the learning outcomes element.This indicates that second objective is achieved.Figure 3 shows percentage of Test 1 question paper generated according to stated learning outcomes.(Note 2)

Advantages of GQS
The system introduced brings several advantages to user compare to manual system.Listed below are the advantages of the system that brings to tourists: i.
Lecturers can generate question randomly by using GQS ii. Save time in preparing paper for the examination.
iii.GQS can helps lecturers to generate the question based on the learning outcomes elements.
iv. Shuffling Algorithms helps randomization process in selecting questions in the database thus preventing duplication and repetition.
v. The question can be added to the bank at any time.
vi. Different sets of questions could be generated without any limitation.
vii.Implementing learning outcomes could classify the question and possible way to evaluate the level of achievement of each course.

Conclusion
The main purpose of this paper is to describe an automatic generator question paper using Shuffling Algorithm for randomization.This system is web-based application system with several features mainly producing unduplicated sets of exam paper.The result indicates the potential evidences of employment such algorithm for this type of system.Our future effort is to employed different types of randomization to see the most effective randomization method for such system.

Table 1 .
Randomization of the question selected using Shuffling Algorithms