Page 83 - Azerbaijan State University of Economics

THE JOURNAL OF ECONOMIC SCIENCES: THEORY AND PRACTICE, V.75, # 2, 2018, pp. 79-84

Element of artificial intelligent can be monitoring system built on the basis of fuzzy
logic, don’t ask hard answer to the question and evaluate answers linguistics. When
building such a monitoring system, you must follow the procedure:

1- Determining how and what question you will ask trainees, and what options for the
answer will be offered to him for thought;

2- Determine their number and meaning of linguistics fuzzy terms which used for the
final assessment of the knowledge of trainees (this number depends on the number
of graduation grade) and to evaluate each current response;

3- Construct a hierarchal evaluation system (output tree), so that at each level of the
hierarchy the number of signs is no more 7±2;

4- Taking into example 1 -3 into account, build the knowledge matrix it is possible to
involve an expert in this field of knowledge, which are the teachers in this specialty.

5- Set the membership function of response level with fuzzy terms, as defined in paragraph
two. And for this you can use the standard membership functions or build a new;

6- To calculate the value of the membership function for each fuzzy terms of each
student answer, and then using the knowledge matrix and output trees, calculate
the value of the membership function of fuzzy terms intermediate feature at each
level the hierarchy level;

7- Using the knowledge matrix, calculate the value of the multidimensional
membership functions for all estimates and their type defined in point 2;
8- Determine the desired solution assessment of the knowledge of the exams of the
trainees. And for this it is necessary to find the maximum value of the membership
function of the answers to this maximum value (see point 7) and to allocate an
estimate, corresponding to this value.

If, however, the values of some of the neighboring membership functions of point 7
matches, that is possible 5 points for action:
A. Correct the original membership function;
B. Correct the knowledge matrix;
C. For estimation use a scale with a large number of gradations;
D. Increase the number of PS questions and add corresponding rows and
comment in the knowledge matrix;
E. To encourage the students, choose a higher rating.
When using the first two options it is advisable to involve experts.

83