A journey from cognitive metrics to cognitive comput

International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN
0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME
60
A JOURNEY FROM COGNITIVE METRICS TO COGNITIVE
COMPUTERS
Dr. Leena Jain1
, Mr. Satinderjit Singh2
1
Associate Professor & Head -MCA, Global Institute of Mgt. &. Emerging Technologies,
Amritsar (India)
2
Ph. D. Research Scholar, Punjab Technical University Jalandhar (India), Associate
Professor, Dept of Computer Applications, GGNIMT Ludhiana.
ABSTRACT
Software cognitive metrics is new field in software engineering with far reaching
potential. The paper profiles the trajectory of software cognitive metrics for last decade and
also enlists some of the concerns and issue related with them. The emergence of Cognitive
informatics (CI) – interplay of computer science, information science, cognitive science, and
intelligence science - with the intent to investigate mechanisms and processes of the brain and
engineering applications arises out is discussed. The key idea of Cognitive computer is also
discussed along with some of path breaking works done or being done under CI. In-fact
Software cognitive metrics looks like just one of many application area of CI.
Keywords: Abstract intelligence, Cognitive computer, Cognitive informatics, Denotational
mathematics, Software cognitive metrics
I. INTRODUCTION
Software’s are inherently complex in nature. From the last few decades it has been the
endeavor of the software industry to find a good measure of the software complexity.
However there are many aspects of software complexity. Fenton distinguishes three kinds of
complexities: computational, psychological and representational [1]. Wang classifies
complexity as computational, symbolic, structural and functional [2]. Yingxu Wang and
Jingqiu Shao proposed a new measure - cognitive functional size (CFS) - of software
complexity based on the cognitive weights [3].
INTERNATIONAL JOURNAL OF ADVANCED RESEARCH IN
ENGINEERING AND TECHNOLOGY (IJARET)
ISSN 0976 - 6480 (Print)
ISSN 0976 - 6499 (Online)
Volume 4, Issue 4, May – June 2013, pp. 60-66
© IAEME: www.iaeme.com/ijaret.asp
Journal Impact Factor (2013): 5.8376 (Calculated by GISI)
www.jifactor.com
IJARET
© I A E M E

61
Earlier measures of software’s complexity typically depended on program size like
counting the number of lines of codes, then some improvement was made by taking into
consideration data flow and module interfaces such as the Halstead’s software metrics and
measure of cyclomatic complexity developed by McCabe became very popular [4], [5], [6].
All these and many other proposed metrics did not capture all aspects of complexity.
The cognitive complexity, as proposed by Wang and Shao provided a new approach
to explain and measure the functional complexity of software as well as the effort in software
design and comprehension [3]. This new approach perceives software functional complexity
as a measure of cognitive complexity for human creative artifacts, which considers the effect
of both internal structures of software and the I/O data objects under processing [3].
This paper studies the development of Software Cognitive metrics in last decade or
so, after they were first proposed by Wang and Shao in the year 2003. In section II a brief
history of development of cognitive metrics in software is profiled. In section III we basically
identify and discuss some of the challenges and issues concerning the next level of
acceptance of cognitive metrics. The doubts and concerns mentioned in section III could not
be properly tackled merely by application of current body of knowledge in computer science.
For this we have to take refuge in emerging trans-disciplinary study of Cognitive Informatics
(CI), which will be introduced in section IV. In-fact we will see that Software cognitive
metrics is just one of many application area of CI. In section V we conclude by pointing out
the exciting new possibilities which are coming out as result of current work being done in
CI. This includes not only the current problem of Software Cognitive metrics and
comprehension, but areas like bridging the gap between Artificial Intelligence (AI) and
Natural Intelligence (NI), Natural Language computing, intelligent search engines, and
computational intelligence. This also includes the exciting possibility to develop a Cognitive
computer- Computer that behaves like human mind with all the features of learning,
reasoning etc.
II. SOFTWARE COGNITIVE METRICS
In the year 2003 Yingxu Wang [3] introduced the concept of cognitive functional
complexity of softwares. In this metrics the BCS basic control structures are assigned
cognitive weights. BCS are the set of fundamental and essential flow control mechanisms that
are used for building logical architecture of software.
TABLE 1
Cognitive Weights of different BCS
BCS Cognitive Weight
Sequence 1
Branch If–Then-Else 2
Case 3
Iterations 3
Function Call 2
Recursion 3
Parallel 4
Interrupt 4

62
In this metrics the total cognitive weight of a component is measured by either adding
the weights of a BCS if they are in series or they are multiplied if they are embedded in
another BCS. The total cognitive weight of a software component, Wc, is defined as the sum
of cognitive weights of its q linear blocks composed in individual BCS’s. Since each block
may consist of m layers of nesting BCS’s, and each layer with n linear BCS’s, the total
cognitive weight, Wc, can be calculated by equation (1).
∑ ∏∑= = =






=
q
j
m
k
n
i
c ikjWcW
1 1 1
),,(( (1)
In this metrics the different BCS are assigned the weights as shown in table 1. These
weights are based on the human effort in comprehending these BCS
The cognitive functional size (CFS) of a basic software component that only consists
of one method, fS , is defined as a product of the sum of inputs and outputs ( oiN / ), and the
total cognitive weight, i.e.:
Coif WNS */= = (No + Ni){ ∑ ∏∑= = =





q
j
m
k
n
i
ikjWc
1 1 1
),,(( } (2)
However [7] have shown that the current existing calculation method of cognitive
metrics can generate different results that are algebraically equivalence.They highlighted the
combinatorial meanings of this calculation method- as shown by Wang [3] and it shows
significant flaw in the measure. Wang’s measure does not take into consideration the data
flow complexity of a component which is not embedded in one another.
TABLE 2
Example of data flow among BCS
for(j=2; j<i; j++)
{ if (i%j= =0)
break;
}
If (i = = j)
printf("t%d",i);
Just to take an example in table 2, Wang’s measure considers the ‘for’ and the ‘if’
structures independently. But the two structures cannot be considered independently, as data
flows from one structure to the other, and in doing so it carries with it some complexity. This
is true because we cannot understand the ‘if’ structure independently without considering the
preceding ‘for’ structure. However in the year 2009, Wang [2] suggested new weights for
various BCS as mentioned in Table 3 below:

63
TABLE 3
Modified Cognitive Weights of different BCS
S.No.
BCS
Description
cognitive
weight
(wi)
1 Sequence 1
2 Branch 3
3 Switch 4
4 For-loop 7
5 Repeat-loop 7
6 While-loop 8
7
Function
call 7
8 Recursion 11
9 Parallel 15
10 Interrupt 22
Although the weights for BCS were changed by Wang, but the method of calculating
overall cognitive complexity of the software remains more or less same. Thus the objection
of [7] still remains. Apart from this there are still some troubling questions about Cognitive
Metrics.
III. ISSUES RELATED WITH COGNITIVE METRICS
The objection of [7], apart, there are some other key concerns regarding Software cognitive
Metrics. Some of the concerns and doubts are as follow:
1) How authentic the weights of BCS are? Can it be statistically verified?
2) Is everybody’s mind work in same way and there is no variation? How does one set of
weights fits all population?
3) What about the variation within the population? To a different population- differing in
age, sex, nationality, computing skills etc. - do these weights hold?
4) Can it be possible – as in case of medicine and social science- to identify various
segments of the population which does not concur with above weights of BCS but have
their own weights?
These and many more questions and doubts of the same category are extremely valid and
important. I am sure in the years to come more and more research in the area of cognitive
metrics will be carried out to fill these concerns or research Gap area.
IV. THE EMERGENCE OF COGNITIVE INFORMATICS
All the concerns and doubts raised in section III are perfectly valid but software
engineering or computer science is not fully equipped to deal with these and other related
issues. The points raised in previous section can only be tackled if we have a proper
understanding of working of human brain; its information processing mechanisms and
about the phenomena of Natural Intelligence (NI). Since the working of human brain was
more or less unknown so there was little scope of advancement in applications (like
Cognitive software metrics) which tries to mimics the working of Human Brain.

64
It was in this backdrop that term Cognitive Informatics (CI) was coined by Wang and
others. It is a separate field of enquiry involving multiple disciplines like computer science,
information science, cognitive science, and intelligence science with the intent to investigate
the phenomena of natural intelligence and the internal information processing mechanisms
and processes of the brain and its engineering applications in cognitive computing. CI was
recognized as separate and important field of enquiry by the scientific community. Even
otherwise CI itself has accumulated a solid body of Knowledge and field is pregnant with
many exciting possibilities. These includes development of highly intelligent systems such
as cognitive knowledge search engines, world-wide wisdom (WWW+), autonomous learning
machines, and cognitive robots.
One key application of CI is idea of Cognitive computer. Actually this is more of an
inspiration and motivation for the researcher working in the field of Cognitive informatics
rather than just being a mere possible application of CI. By Cognitive Computer we mean an
intelligent computer which has a capacity of human brain; that it can perceive, learn and
reason and is basically a knowledge processor rather than a data processor which a
conventional computer is. The current problem of Software cognitive metrics will be just a
small little application in Cognitive computer and some of the concerns raised in section III
should probably be tackle-able in it as well. Also the distant dream of computer linguists of
Computing with Natural Language (CNL) should now seem possible through Cognitive
computers [8].
V. THE PILLARS OF COGNITIVE RESEARCH
The previous section introduces the Field of Cognitive Informatics (CI). CI borrows
the work done in so many diverse fields, seals it under one roof and enriches these works by
developing a common theoretical frame work and supporting mathematics for developing the
common understanding of Human Brain and subsequent computing application coming out
of it. In this context some of path breaking works done or being done under CI can be
basically classified in following categories.
A. Theoretical framework of CI
Since its advent a decade ago, CI has developed a phenomenal body of work in
building up theory of how human Brain works. This includes Information-Matter-Energy
(IME) model, the Layered Reference Model of the Brain (LRMB) , the Object-Attribute-
Relation (OAR) model of information/knowledge representation in the brain [9], [10].
Particularly the development of Layered Reference Model of the Brain (LRMB) is of
extreme importance and is really a milestone. It identifies 43 cognitive processes in 7 layers.
All human efforts or activities are supposed to be permutation and combination of these 43
cognitive functions. Why LRMB is an important development? Because with the
identification of basic cognitive processes, it has become possible, first time human history,
to have computer representation of basic human features of consciousness, emotion,
intelligence etc. And it is generally believed that the study on integration of consciousness,
emotion and intelligence is a good stepping stone to cognitive computing.

65
B. Common concepts for all types of Intelligence
Even after the 57 years of formal research in Artificial Intelligence (AI), till date we
have still not produced a machine which can pass the Turing test [11]. It has become more
and more clear that modern computing – impressive as it may be- is nowhere near human
intelligence level. As noted by Zadeh, human are remarkable in two ways at least. First of all
they have a remarkable ability to take decision in environment of uncertainty, imprecision,
and incomplete information [11]. And the ability to perform task without any measurement
and computation is also something which is different from computers. Clearly natural
intelligence (NI) can not be replaced by Artificial Intelligence (AI), until the very process of
internal information processing of brain is known properly.
In this context the development of novel theory of Abstract intelligence (αI) is quite a
remarkable achievement in its own right [12]. It is an attempt to bring both form of
intelligence –Artificial and Natural- under one common umbrella and then fits it to neural,
cognitive, functional and logical form. The paradigms of Abstract intelligence are natural,
artificial, machine-able, and computational intelligence. The study of Abstract intelligence
(αI) aims at revealing the basic mechanisms of different forms of intelligence.
C. Development of Supporting Mathematics for CI
The development of any scientific field normally goes hand in hand with
mathematical representation of concepts underlying its theory. One of the endeavor of CI
since its inception is to develop parallel mathematical structure able to depict complex
concepts and idea inherent in its theory. And it is more imperative for field like CI, which is
multi disciplinary in nature. The absence of proper Mathematical support would lead us to
days when CI was not conceived. Different fields were working in their own smaller
domain areas- perusing their small goals- but their collective effect were missing. Large
category of DMs have been created and developed in CI such as real-time process algebra
(RTPA) [13], concept algebra, system algebra [14], granular algebra, and visual semantic
algebra (VSA).
VI. CONCLUSION
The formal interest in Cognitive informatics started in the year 2002, when IEEE
recognized it as a separate field and lends its name and recognition to International
Conference on Cognitive Informatics (ICCI). Since then the interest in the field has grown
tremendously. Yingxu Wang has played a key role in the development of CI. He is the
founder and steering committee chair of the annual IEEE International Conference on
Cognitive Informatics and Cognitive Computing (ICCI*CC). It was coincidence that one of
the first problems he dealt in 2002-03 was to look towards software in terms of cognitive
complexity and introduced the concept of BCS along with their weights. Or perhaps it was
just not the coincidence. Whatever might be the case; but it is amazing that something which
started with measuring the effort to comprehend the software cognitively has expanded itself
to become one of the most exciting, emerging and trans-disciplinary field in applied science.
The paper reflects this journey.

66
REFERENCES
[1] N.Fenton,(1994),”Software Measurement: a Necessary Scientific Basis”, IEEE Trans.
Software Eng., Vol. 20, No. 3, pp. 199-206.
[2] Y. Wang,(2009) “On the Cognitive Complexity of Software and its Quantification and
Formal measurement.” International Journal of Software Science and Computational
Intelligence, Vol. 1, No. 2, , pp. 31-53.
[3] Y. Wang. and J. Shao (2003), "Measurement of the Cognitive Functional Complexity of
Software", Proceedings of the 2nd IEEE International Conference on Cognitive Informatics,
2003, pp. 67-74
[4] A.J. Albrecht and J. E. Gaffney (1983),“Software Function, Source Lines of code, and
Development Effort Prediction: A Software Science Validation”, IEEE Transactions on
Software Engineering, Vol.9, No.6, , pp. 639-648.
[5] M. H. Halstead,(1996) , “Elements of Software Science” (Operating and programming
systems series), Elsevier Science Inc. New York, Edition 1977, pp. 68-96.
[6] T. H. McCabe, (1976) “A Complexity Measure”, IEEE Trans. Software Eng. SE – Vol.2,
No. 6, pp. 308-320.
[7] B. Auprasert, and Y. Limpiyakorn,(2008) , “Underlying Cognitive Complexity Measure
Computation with Combinatorial Rules”, Proceedings of world academy of science,
engineering and technology Vol. 35, 2008, pp 432-437.
[8] L.A. Zadeh, (1999), “From Computing with Numbers to Computing with Words – from
Manipulation of Measurements to manipulation of Perception”, IEEE Trans. on Circuits and
Systems I, 45(1), 1999, pp. 105-119.
[9] Y. Wang, S. Patel, and D. Patel (2006) , “A Layered Reference Model of the Brain
(LRMB)”, IEEE Trans. on Systems, Man, and Cybernetics (C), 36(2), , pp. 124-133.
[10] Y. Wang(2007), “The OAR Model of Neural Informatics for Internal Knowledge
Representation in the Brain“, International Journal of Cognitive Informatics and Natural
Intelligence, 1(3), 2007 pp., 64-75.
[11] L.A. Zadeh, (2008a), “Toward Human Level Machine Intelligence - Is It Achievable? The
Need for a Paradigm Shift”, Computational Intelligence Magazine, IEEE , 3(3), 2008, pp. 11
– 22.
[12] Y. Wang(2008b), “On Contemporary Denotational Mathematics for Computational
Intelligence”, Transactions of Computational Science, 2, Springer, pp. 6-29.
[13] Y. Wang(2008c) ,” RTPA: A Denotational Mathematics for Manipulating Intelligent and
Computational Behaviors”, International Journal of Cognitive Informatics and Natural
Intelligence, 2(2), 2008 , pp. 44-62.
[14] Y. Wang, L.A. Zadeh, and Y. Yao (2009b), “On the System Algebra Foundations for
Granular Computing” , International Journal of Software Science and Computational
Intelligence, (1), 2009, pp. 1-17.
[15] J.Emi Retna, Greeshma Varghese, Merlin Soosaiya and Sumy Joseph, “A Study on Quality
Parameters of Software and the Metrics for Evaluation”, International journal of Computer
Engineering & Technology (IJCET), Volume 1, Issue 1, 2010, pp. 235 - 249, ISSN Print:
0976 – 6367, ISSN Online: 0976 – 6375.
[16] Mohammed Sirajuddin, Dr D. Rajya Lakshmi, Dr Syed Abdul Sattar and Nafisur Rahman,
“Fuzzy Logic — The Fascinating Logic Behind Artificial Computational Intelligence”
International Journal of Advanced Research in Engineering & Technology (IJARET),
Volume 4, Issue 3, 2013, pp. 280 - 285, ISSN Print: 0976-6480, ISSN Online: 0976-6499.

A journey from cognitive metrics to cognitive comput

Recommended

Recommended

More Related Content

What's hot

What's hot (16)

Viewers also liked

Viewers also liked (19)

Similar to A journey from cognitive metrics to cognitive comput

Similar to A journey from cognitive metrics to cognitive comput (20)

More from IAEME Publication

More from IAEME Publication (20)

Recently uploaded

Recently uploaded (20)

A journey from cognitive metrics to cognitive comput