Oleg Kotenko

Speciality: «System programming»
Group: SP-01m
email: kotenko@istil.com.ua
Master's thesis: «Development of resources and methods of automation of question creation process in distant knowledge testing systems»
Project crew: Zinchenko Y.E.

Oleg Kotenko

ru | ukr

Donetsk national technical university

Masters DonNTU portal

Abstract of master's thesis

Introduction

At present, distant knowledge testing systems, finding a use in many science and engineering fields, became more popular.

Now some quantity of testing system exists, and many of them have the following imperfections:

Also, because of growing popularity of testing systems, new disadvantage appeared: lack of automation of question creation process. It's especially noticeable in case of great quantity of tasks, for example, when preparing system for entrant testing or when examining the professional appropriateness in enterprises.

Development of resources and methods of automation of question creation process in testing systems is a topic of this work.

Automation essence

The problem is to find a way to generate a great number of the same kind of questions, using as minimal efforts and working time as possible. This problem solving is possible in case of creation of technical question database, for example, when testing university students.

It is difficult to automatically create non-textual questions, for example, questions that include some graphics or graphical at all.

Methods of automation

Process of creation of questions which have textual part as main semantic load is easier to automate.

In this case questions are generated using created template following way:

Third stage is the most labor-intensive. Using this method of question creation process automation it is possible to get many questions that have no sense, so there's a parallel task to find a way to avoid such situation or to eliminate it without considerable efforts.

Interrelations between changing elements can be clearly pictured as directional graph, with elements as vertexes and relations between elements, which determine finite set of changing elements in resulting question, as verges.

Simplification of nonsensical question elimination process can be reachable using GUI, added to questions constructor, which can clearly represent interrelations of changing elements and easily edit them. However, useless questions are probable, but their quantity lessens noticeably and they can be removed later manually without efforts.

Using this template, creation process comes to mere sorting out combinations of changing elements an inserting every set to textual template.