Intelligence quotient of information systems

DEFINITION

edit

One of the important features of human intelligence is the ability to make right decisions in a situation of inaccurate and incomplete information. Construction of models, which are similar to human thinking and their use in a computer is an important and topical issue.

Development of the intelligent systems within the last years fetch to do such systems which have ability to think, to learn from experience, and to make rational decisions without human intervention. Prof. Zadeh has coined the term 'MIQ' (Machine Intelligence Quotient) to describe the measure of intelligence of man-made systems. In this perspective, an intelligent system can be defined as a system that has a high MIQ.

Intelligence quotient on one information system (ISIQ) is the proportion between obtained level of intelligence and the average level of intelligence of an age group, which it refer to multiplied by 100. [1]

Defining of ISIQ is accompanied with filling in a test, in which are included questions from all kinds of intelligence and the distribution between them all is equal [2]:

  • Il linguistic intelligence at information system (IS)
∙ amounts of a word;
∙ reasonable functions – methods for identification of meaning on a word, inquiry;
∙ functions of syntax, methods for identification of derivative words, dialects.
  • Ig spatial intelligence at IS
∙ association functions – methods referring to the specific groups, category;
∙ visual/view possibility;
∙ function to distinguish of images, objects;
  • Im musical intelligence at IS
∙ voices functions;
∙ phonetically functions, identification of a voice, a sound.
  • Id motor intelligence at IS
∙ pability for performance on movements in her separate controlled objects.
  • Ie emotional intelligence at IS
∙ Functions for the character recognition;
∙ Functions for recognition of the extremes, inaccuracy.
  • Ilm logical-mathematical intelligence at IS
∙ logic functions;
∙ composition from mathematical operations, formulas.
  • Ietcother intelligence at IS
∙ areas of the knowledge;
∙ actuality;
∙ time for a recognition meaning on inquiry;
∙ time for reception of the answer.

After answering half of the questions of the test an average level of intelligence on information systems – 100 (mISIQ=100) is determined. Besides when estimating the intelligence of the intelligent information system - AIIS must be taken into account the age corresponds to the interval from the moment of functioning of the system as an intelligent one to the moment of applying the test.

Definition: The average level of intelligence in IS’s (mISIQ) is 100 for different age groups. Between the different age groups is included the correction quotient of age (Кaiis), which specifies mISIQ.

Determining this correction quotient is performed further through comparison of the results of a test for one age group with the results for all other age groups.

For ISIQ is valid the formula:

 

, where Bt is count of positive answered from IS questions, Kaiis correction quotient of the age, mISIQ is average level of intelligence in IS and Ba is count of questions in test.

Example: It is necessary to determine ISIQ in IS when carrying out a test, which includes 36 questions, the system answered positive to 25. To include Кaiis it is necessary that tests are conducted over several age groups with more than one IS in a group. As it is impossible we will accept a common age group. It follows from this that mISIQ = 100 if 18 questions (50%) from the test are answered. For each correct answer 5.56 is added to ISIQ. In this case ISIQ = 25×5.56=139.

EXAMPLE FOR TESTING OF INTELLIGENCE QUOTIENT ON ONE INFORMATION SYSTEM

edit

Test for determining the IQ of information system, consisting of 28 questions with equal weights (1), each successive group of four questions meets the consecutive intelligence:

# Kind Iint Question Answer
1 Il Recognize correct word "student", "studet"  Y Done
2 Il Reaction systems of arbitrary word "faculty"  Y Done
3 Il Face typo between the words "engineer" and "engeneer".  Y Done
4 Il Face saved word on any natural language - "good".  Y Done
5 Ig Submit a response of the same requesting, which set query.  Y Done
6 Ig Recognition of the professional interests of the applicant.  Y Done
7 Ig Detection of visual change in the same query.  Y Done
8 Ig Recognition of physical presence of the applicant, without direct interaction with it.  Y Done
9 Im Recognition of correctly pronounced the word "Christo", "Risto".
10 Im Speech reaction.  Y Done
11 Im Recognize to saved wrong words, which are pronounced similarly - "Insurance", "Insurence".  Y Done
12 Im Recognize the meaning of the word with different pronunciation "potòka", "pòtoka".
13 Id Responding to written task - Generate a printout of a list of personal data of staff.  Y Done
14 Id Responding to written task - Delivery of results of the teacher Ivan Ivanov (not electronically).
15 Id Responding to a written assignment - "Archive on the outcome of inquiry to information medium".  Y Done
16 Id Responding to a written assignment - "Replace toner cartridge printer".
17 Ie Recognition of non-operational of the applicant.  Y Done
18 Ie Recognition of the words "big", "small".  Y Done
19 Ie Different response to a written assignment "Please for list of students of specialty Sociology" and "Generate a list of students of Sociology".
20 Ie Recognizing the desire of the applicant.  Y Done
21 Ilm Distinguishing numbers recorded in different numeral system - "1", "12".  Y Done
22 Ilm Using the numeric equivalent of the value delivered by voice.  Y Done
23 Ilm Using the numeric equivalent of the recorded number of text - "two".  Y Done
24 Ilm Solving the expression specified in writing "1+5" and generate text result.  Y Done
25 Ietc Recognition of words from different subject area - "Informatics", "Literature".
26 Ietc Generating a response to job within the time specified by the upcoming.  Y Done
27 Ietc Generating the correct response to an assignment within the time specified by the user.
28 Ietc Generating the correct response task.  Y Done

That test includes 28 questions, the system answered positive to 21. To include Кaiis it is necessary that tests are conducted over several age groups with more than one IS in a group. As it is impossible we will accept a common age group. It follows from this that mISIQ = 100 if 14 questions (50%) from the test are answered. For each correct answer 7.14 is added to ISIQ. In this case ISIQ = 21×7.14=150.

CONCLUSION

edit

Definition on term intelligence quotient on information system suppose put of one new class among information systems – intelligent information systems (IIS).

Determination of degree of intelligence on information systems by means of introducing on average level of intelligence on IS – mISIQ and intelligence quotient on IS - ISIQ it is one method for estimation on them intelligence.

Inclusion of the correction quotient Кaiis on age on the further estimation will be bring additional exact and precision at determination on the ISIQ.

All this, will be, accelerate advance on that class IS, as well as establishment of the competitive IIS according to step of intelligence with purpose of use them in society. On the other hand their advance and improvement will be, contribute to make on information service with the most high quality and accuracy, with methods and the reaction inherent to the person, as in any cases – and better.

REFERENCES

edit
  1. Rashidov A., T. Nenov, Estimation of the degree of Intelligence of information systems, Journal of Automation and Informatics, 2006, No 2, pp. 37-40, ISSN 0861-7562, In Bulgarian
  2. Rashidov A., Estimation of the degree of Intelligence of Database Management Systems, CEEPUS WEB Summer School 2007– Information Systems in Automation, Maribor, Slovenia, 2007, in CD-ROM. ISBN 978-961-248-054-7/ COBISS.SI-ID 59395585.