An enhancement of ant colony algorithm applied in automatic exam generation / Cabisora, Angelica B. and Calleja, Isabel D. 6

Undergraduate Thesis: (BSCS major in Computer Studies) - Pamantasan ng Lungsod ng Maynila, 2016. 56

5

ABSTRACT: 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 challenging, tendious 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. Usually the instructors keeping 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. Traditional method of test paper generation is also inefficient and has low success. So, to improve the process of creation of test paper generation, the ant colony algorithm is applied to it. Ant algorithms were inspired by the observation of real ant colonies. Ants are social insects, that is, insects that live in colonies and whose behavior is directed more to the survival of the colony as a whole tan to that of a single individual component of the colony. Scientists have found that the complex behavior of ants can provide models for solving difficult combinatorial optimization problems. Ant Colony algorithm was introduced by Marco Dorigo in his Ph.D. thesis in 1992 and was called Ant System (AS). Ant System is the result of a research on computational intelligence approaches to combinational optimization at Politecnico di Milano in collaboration with Alberto Colorni and Vittorio Maniezzo. Initially, the algorithm was applied to the Travelling Salesman Problem (TSP), and to the Quadratic Assignment Problem (QAP), Network Model Problem and Vehicle Routing. Ant Colony shows great performance with the ill-structed problems. Moreover, Ant Colony algorithm is extremely important in local search to obtain good results. Ant Colony algorithm is a class of optimization algorithms for finding optimal paths which is inspired by the food search behavior of ants and their ability in finding optimum paths by using special chemical pheromone to communicate between colonies. In real ant case, when a food is located, ants roams randomly from their colony to food. Upon finding food, they will deposit a chemical called pheromone on their way back to their home. So, when other ants come across the pheromones, they are likely to follow the path with a certain probability. If they do, they populate the path and leave pheromones too which will make the pheromone concentration stronger. Thus when one ant finds a good path from the colony to the food source, other ants ,may likely to follow the same path and positive feedback eventually leads to all the ants following the same path.

5