We are IntechOpen,

the world’s leading publisher of

Open Access books
Built by scientists, for scientists

6,100
Open access books available
149,000
International authors and editors
185M
Downloads

Our authors are among the

154
Countries delivered to
TOP 1%
most cited scientists
12.2%
Contributors from top 500 universities

Selection of our books indexed in the Book Citation Index

in Web of Science™ Core Collection (BKCI)

Interested in publishing with us?

Contact [email protected]
Numbers displayed above are based on latest data collected.
For more information visit www.intechopen.com
Chapter

Application of Fuzzy Expert

Systems in IT Project Management
Oleksii Dudnyk and Zoia Sokolovska

Abstract

The available statistics show the growing influence of the IT market on the world
economy over the last decade. According to expert information, this situation will
continue, despite the IT sector’s economic crises, uneven development, and periodic
fluctuations. The need to involve fuzzy expert systems (ES) in the IT field is stated,
based on the high uncertainty level due to specifics of IT project management. The
hypothesis of embedding ES in an IT company’s business process management to
increase the efficiency of operational and strategic decisions is tested. The structure of
ES is offered, built on the basis of fuzzy logic using a combined model of the semantic
network and implication rules. The operation of the system is demonstrated in the
example of managing an IT company’s current business processes to maximize its
profits. Comparing the conclusions of the ES with the historical decisions of a real
company demonstrates the feasibility of implementing the ES. The operation of the
developed ES, using the knowledge base formed on the basis of 30 Ukrainian IT
companies, confirmed the effectiveness of its use as a tool to support management
decisions and increase the IT sector’s financial performance.

Keywords: IT market, IT company, expert system, fuzzy logic, IT business processes,

management decisions, project management

1. Introduction

One of the distinguishing features of recent years has been the exponential growth
of digital data aggregation. This is accompanied by the expansion of big data analytics,
artificial intelligence, cloud computing, and digital platforms. As more devices access
the Internet and the number of people using digital services grows, the role of digital
data and technology is becoming more widespread, and the digital economy is evolv-
ing at a breakneck pace [1]. The information and communication technology (ICT)
industry is at the heart of much of this activity, supporting the digital economy and
serving as a reliable measure of its effectiveness.
ICT plays a crucial role in the economy not only as a source of potential income but
also as a vector of cross-growth, making profound changes in various sectors of the
economy. Technologies such as the Internet of Things, robotics, artificial intelligence,
cloud computing, big data analysis, 3D printing, and many others are already
changing the way businesses design, produce and provide services.

1
Project Management - New Trends and Applications

Despite how important IT projects are to all aspects of the modern world, their
management is not an ideal process. A recent survey conducted by Standish Group
showed that 19% of over 50,000 software projects are failed and never completed [2].
At the same time, over the previous 10 years, the percentage of failures varied from
17–22%, which indicates the regularity of this problem in software engineering [3]. In
the context of numerous crisis phenomena of the world economy, against the back-
ground of these trends, the importance of the main subjects in the IT industry (prod-
uct and outsourcing IT firms) and tools to improve the efficiency of their operational
business processes is growing. One of the innovative directions is the use of technol-
ogies like ES in the management processes to better plan and execute project devel-
opment. Although the use of expert systems in economics and management is not
entirely new, the dynamics of research and application of this artificial intelligence
apparatus over the past 20–40 years have undergone significant changes and fluctua-
tions—from discovery and active development to a significant decline, and again to
the current trend of scientific recovery and the practical interest of specialists in
various fields.
Accordingly, the tasks of the analysis of the history of development and use of
expert systems in IT project management are set. The expediency and prospects of
using the ES in the management of business processes in IT companies as innovative
management methods that take into account the vagueness and uncertainty of the
information environment of the facilities are proven. The positive impact of the
introduction of intelligent technologies, including fuzzy ones, on the management
processes of IT companies is demonstrated, which is reflected in the main financial
indicators of their operation.
The article is organized as follows. Section 2 confirms the relevance of embedding
expert systems in the overall business process management of IT companies. A retro-
spective of ES research and examples of applications of specific applications in various
fields of economics and management are presented. Section 3 discusses the working
hypotheses of the study; the ES architecture developed on the basis of fuzzy logic is
offered. Section 4 provides a statement of the simulated situation, which is used for
expert consultations using the appropriate knowledge base; an example of forming a
knowledge base is described based on information provided by 30 IT companies of
Ukraine, as well as using the expertise of agile specialists; the conclusions of the fuzzy
expert system obtained as a result of its implementation (expert consultations) and
historical decisions made by a functioning IT company are presented. Section 5 pro-
vides a comparative analysis of the results of the expert system inference conclusions
and the real conclusions of a functioning IT company. Section 6 is devoted to outlining
potential ways of further research to determine the useful consequences of the imple-
mentation of the proposed mathematical apparatus.

2. Historical analysis and overview

2.1 IT project management research analysis

The characteristic features of a software project are a large amount of research and
development work, high uncertainty in the type, timing and cost of work, significant
risks, and high costs. On top of that software development is associated with a high
degree of complexity in a constantly changing environment. Thus, software projects
demand effective management using innovative approaches.
2
Application of Fuzzy Expert Systems in IT Project Management
DOI: http://dx.doi.org/10.5772/intechopen.102439

Problems of IT project management are researched in many literature sources.

Tam et al. in Ref. [4] conducted a survey of 216 agile professionals and identified the
main human factors for the success of agile software development projects. In Ref. [5]
Fink and Pinchovski presented an empirical study of the bias of the decision to save
time in a software development project by increasing the speed of development and
proposed methods to combat this bias. In Ref. [6] Hoffmann et al. presented the
principles of designing strategically consistent and effective, but flexible portfolios of
IT projects. Einhorn et al. in Ref. [7] consider the importance of taking into account
the business case throughout the project life cycle and also provides valuable infor-
mation on ways to avoid common mistakes and achieve the planned strategic benefits
of the project. Lin et al. explore in Ref. [8] how to improve the integration of knowl-
edge by actively addressing the problem of project uncertainty and proposes different
management regimes, taking into account the types of uncertainty. In Ref. [9]
Bjorvatn and Wald conducted an empirical study of the relationship between project
complexity and management efficiency and determined the crucial importance of
absorption capacity at the team level. Keil et al. in Ref. [10] research the impact of
project management constraints on the possible escalation of software projects. In
their work, Tavares et al. [11] analyzed different risk management strategies carried
out in Scrum software projects and developed a novel risk management framework.
One of the key issues faced by both software developers and customers is to
consider the degree of risk inherent in the various stages of IT project deployment. In
[12–17] the classification of risks, their sources—related to incorrect determination of
project volumes [13], excess of project costs [14], errors in budget planning [15, 16],
deviations from deadlines performance [17].
Several works are devoted to automating some parts of IT project management
activities. Alba and Chicano in Ref. [18] applied genetic algorithms to optimally
allocate resources for IT projects. Uzzafer in Ref. [19] presented a simulation model
for strategic IT project management. The results of the simulation determine the
budget and schedule required for a project.
In the process of preparing and implementing a software project, the manager is
forced to make decisions in conditions of uncertainty, based on incomplete or inaccu-
rate information about the current state and prospects of the project’s development. It
is possible to improve the quality of the decisions made by integrating an intelligent
component—an expert system—into the decision-making process. However, the use
of expert systems in the process of IT project management is almost beyond the
attention of researchers.

2.2 Expert system research and usage in IT project management

An expert system is a computer program that, based on the rules laid down in its
knowledge base, can give reasonable advice and suggest a solution to a problem. The
use of the expert system as a decision support tool is justified for solving problems that
cannot be solved based on analytical calculations.
Work on the creation of expert systems began in the early 1950s by Newell et al.
[20], who developed a common problem solver for solving problems of elementary
logic, proving theorems, and playing chess. This approach underestimated the role of
specific knowledge in reasoning. Aware of the possibilities, research has been
conducted in more specific areas of knowledge, such as medicine and chemistry. The
first expert system was developed in 1965 by Feigenbaum et al. [21] and was intended
for the analysis of chemical compounds. Since then, the range of applications of expert
3
Project Management - New Trends and Applications

systems to industrial and commercial problems has become so widespread that they
have become one of the most successful commercial areas of artificial intelligence.
Some examples of the ES use in various areas of business are discussed below.
One example of the ES application in management is the software application
Business Insight [22]—an expert system to support decision-making and strategic
planning. Insight business presents the user with opportunities for strategic analysis,
business monitoring; identification of key factors influencing business success;
strengths and weaknesses of the business; obtaining the results of forecasts for the
implementation of various business strategies. Starting with the user’s answer to the
questions asked by the system during the introduction, it can conduct a number of
analyzes, providing the user with practical understanding and advice on his business
and marketing strategies. The system also shows the progress of its logic for each
comment or recommendation it makes.
In Ref. [23] Rao et al. present an expert system called PAT (productivity assess-
ment technology), which provides a comprehensive analysis of project effectiveness.
PAT uses the same logical process as a specialist in this field would identify the causes
of good or bad performance. The proposed system also recommends corrective action
and provides the user with explanations or justifications for the results.
When discussing the pros and cons of expert systems, most researchers focus on
the list of advantages of expert systems and pay less attention to the disadvantages. In
Ref. [24] Zarandi et al. focus on the weaknesses of expert systems, namely:

• Lack of ability and flexibility to adapt to changes in the environment.

• Lack of ability to generate a creative answer when there is no answer.

• Lack of ability to summarize their knowledge using an analogy.

• Impossibility to learn: usually, expert systems do not have the opportunity to

learn from experience. Many expert systems cannot automatically change their
knowledge base, nor adjust existing rules or add new ones.

One of the methods of combating these shortcomings is a combination of expert

systems with methods and techniques of machine learning and artificial intelligence.
For example, fuzzy logic can be used to manage uncertainty in expert systems and
solve problems that cannot be effectively solved by conventional methods [25]. The
main purpose of fuzzy expert systems is to use human knowledge to process uncertain
and ambiguous data. Fuzzy expert systems have a history of use in various fields, in
particular in economics and IT.
One of the potential areas of ES and fuzzy ES applications is IT projects manage-
ment and their inherent business processes. Information technology projects are par-
ticularly prone to failures due to their specific characteristics, such as the lack of clear
constraints, the complexity and abstractness of tasks, and the extremely rapid pace of
technological progress. These factors increase the uncertainty, inaccuracy, and sub-
jectivity in information technology projects and require the search for new manage-
ment methods. Here are some examples of the ES application in the field of
information technology.
The work of Dufner et al. [26] discusses the PMA expert system (Project Manage-
ment Advisor), which can improve control over the IT project by evaluating the
proposed project plan, identifying anomalies, and providing guidance for correction.
4
Application of Fuzzy Expert Systems in IT Project Management
DOI: http://dx.doi.org/10.5772/intechopen.102439

The PMA was developed as part of the CyberCollaboratory [27], built to facilitate
collaborative design work. The PMA was approved by industry experts involved in the
knowledge acquisition process and evaluated on 11 realistic project plans. The results
showed a clear ability to detect anomalies in the project plan. The PMA also provided
explanations and suggested corrective action.
Truicǎ and Barnoschi present in Ref. [28] an expert system for recruiting IT
specialists, which helps the human resources department to perform the recruitment
of qualified specialists, assessing their skills, and offering advice on appointments. The
system is designed to work in the field of information technology. Checking the
accuracy of the system showed that the system selected the same three best job
candidates as the expert person.
The work of Rodríguez et al. in Ref. [29] proposes a new method of risk assessment
for the analysis of projects in the field of IT. The proposed method is based on a
combination of a fuzzy process of analytical hierarchy and a system of fuzzy infer-
ence, benefiting from their advantages and minimizing their disadvantages. The pro-
posed model takes into account different levels of uncertainty, the relationship
between groups of risk factors, and the possibility of adding or suppressing variations
without losing consistency with previous estimates. A case study of three actual IT
projects showed the suitability and consistency of the proposed method results.
However, despite the fact that the field of information technology is very promis-
ing for the use of fuzzy expert systems, a review of the literature shows extremely
little use of this apparatus in this area.

3. Fuzzy expert system application design

Historical review of the development and use of expert systems show both the
prospects of the direction and the lack of attention to it. Next, we will focus on the
architecture, methodological platform, and usage of the developed application—a
fuzzy expert system. We will prove the possibilities and efficiency of its use in the
business process management of a typical IT company.
Further considerations are based on the following hypotheses:

1. Expert systems, as the apparatus of artificial intelligence, can be used as an

innovative method of managing economic processes and systems by taking into
account the various informal influences of their uncertain environment.

2. Expert systems should be built into the overall business process management of
IT companies, which has a positive effect on the effectiveness of operational and
strategic management decisions.

3. The use of the expert system applications, based on fuzzy logic, in the process of
managing the IT company business processes, increases the level of the main
financial indicators, in particular, net profit.

Classical ES architecture based on inference: Database with initial information

necessary to get an output; base of facts for the preservation of intermediate results;
knowledge base with information on inference process through the knowledge base
and fact base, and the core of inference (see Figure 1). The most important compo-
nents that make sense to explore are the knowledge base and the core inference.
5
Project Management - New Trends and Applications

Figure 1.
The classic mechanism of inference.

One of the disadvantages of the classical inference architecture is the constant need
to access the database to obtain the necessary information to calculate and maintain
the database of facts in the current state. The second disadvantage creates unnecessary
questions about the structure and ways to maintain a database of facts, which can be
physically part of the knowledge base, which is illogical, or part of the database, which
increases the load on the database.
It is possible to get rid of both shortcomings by reviewing how to work with the
knowledge base and ways to maintain it. Instead of using crisp inference rules, there is
a possibility to use a combination of fuzzy inference rules and simplified linguistic
variables, which will be described below. This structure of the knowledge base allows
getting rid of the intermediate facts database, which was closely related to the need to
store intermediate information in order to have permanent access to the database.

3.1 Knowledge base structure

Before moving on to the use of semantic networks for fuzzy inference core,
consider the second important component—the knowledge base. In the terminology
of fuzzy expert systems, the knowledge base is a set of inference rules and linguistic
variables, on the basis of which the mechanisms of direct and inverse inference work.
A production rule can be defined as an IF-THEN structure that links information or
facts in an IF part to certain actions or information in a THEN part. Thus, the base of
production rules can be composed of an unlimited number of rules of the form:

IF ðX ¼ AÞ THEN ðY ¼ CÞ, (1)

X, Y are linguistic variables;A, C are fuzzy linguistic equivalents of some crisp

value associated with the corresponding linguistic variable.
The second part of the knowledge base is a set of linguistic variables that consists
of an unlimited number of variables of the form:
6
Application of Fuzzy Expert Systems in IT Project Management
DOI: http://dx.doi.org/10.5772/intechopen.102439

X : Initial : ½A : Trapezoidal : ðA1, A2, A3, A4ÞjB : Trapezoidal : ðB1, B2, B3, B4ފ, (2)

X is a linguistic variable;Initial/Derivative indicates that linguistic variable value

will be taken, for example, from a database or the value of which will be derived in the
process of working with the knowledge base;A: Trapezoidal:(A1,A2,A3,A4) and B:
Trapezoidal:(B1,B2,B3,B4) are the fuzzy membership functions;A, B are sets of values
for a linguistic variable X;Trapezoidal indicates a trapezoidal type of membership
function used to describe the values of a linguistic variable;(A1,A2,A3,A4), (B1,B2,B3,
B4) are crisp values behind the fuzzy values of A and B, respectively.
This structure of the knowledge base has several very important features. First, it is
possible to expand and reuse. The knowledge base can be expanded with new knowl-
edge in the transition from specialist to specialist. Second, it is a potential opportunity
to combine a knowledge base and a database to simplify the creation of a knowledge
base based on production rule templates and linguistic variables. In this case, it will be
possible to use rule templates instead of the structures of inference rules and linguistic
variables described earlier, which will significantly speed up the process of creating a
typical content of the knowledge base.

3.2 Inference core: SNePS

Next, we consider the process of fuzzy inference, namely the use of semantic
networks for fuzzy inference. Semantic networks have been developed to present
knowledge of an intelligent system that uses natural language. For the fuzzy inference
problem, it was decided to use the SNePS semantic network processing system from
the study of Shapiro and Rapaport [30], which is a denoted directional graph in which
nodes represent concepts and arcs represent binary relationships between concepts. A
feature of the SNePS semantic network is access to the database once to obtain the
initial data. The inference rule can be represented in a graph through the nodes of the
rule itself, the formulas of the input and output arguments, as well as the arcs that pass
from the nodes of the rule to the nodes of the arguments. We should not forget about
the connection between the rules due to the inclusion of linguistic variables from the
right or left part of one rule in the right or left part of another rule. If the left part of
one rule occurs in the right part of the second, the second rule is called the predeces-
sor. Otherwise—a follower. Consider the following example and semantic network for
this set of rules (see Figure 2):
R1: IF (A1) THEN (B1)
R2: IF (A2 AND A3) THEN (B2)
R3: IF (B1 AND B2) THEN (C1)
Rule R1 is an equivalence rule, which means the following—if the predecessor is
TRUE, then the successor also becomes TRUE. Rules R2 and R3 are general inference
rules, which means that each predecessor must be TRUE for the followers to be TRUE.
Inference graphs were proposed and developed by Schlegel and Shapiro in Ref. [31]
as extensions of propositional graphs. An inference graph is a graph of reasoning that is
capable of inverse, direct, and bidirectional inference. It can support parallel processing
for reasoning using inference logic. Inference graphs modify propositional graphs by
adding channels between nodes along possible inference paths. Channels carry priority
messages to transmit new information from one node to another. Message priorities
affect the order in which tasks are performed so that messages are executed closer to the
inference output and the inappropriate output tasks are canceled. A rule node is capable
of performing inference operations using a set of rules known as Rule Usage
7
Project Management - New Trends and Applications

Figure 2.
Semantic network for a set of rules.

Information (RUI) [32]. RUIs contain information about which predecessors are true
TRUE or FALSE, as well as information that explains how these values were derived.
When a new RUI is created, it is combined with a set of existing ones. The resulting
combination is used to determine whether the inference rule node can be used again.
For fuzzy inference in the developed system, only the direct inference process is
used. Therefore, the structure of the inference graphs can be simplified:

1. There is no need for parallel reasoning, so there is no need for priorities for
messages.

2. Instead of RUI, it is possible to use the simple status of the rule, which is updated
when a new message is received. The status contains the result of counting the
statuses of all rule predecessors.

Figure 3 represents a graph from Figure 2 with the corresponding channels.

Channels allow predecessors to report rule nodes when it was calculated and also allow
rule nodes to report that they have been calculated. With this use of the semantic
network, the dependence on the initial data in the database is reversed so that the core
of the inference mechanism, that is, the semantic network, does not need to

Figure 3.
Semantic network for a set of rules with channels.

8
Application of Fuzzy Expert Systems in IT Project Management
DOI: http://dx.doi.org/10.5772/intechopen.102439

Figure 4.
Modified inference mechanism.

constantly query the data, but only rely on existing initial data and the potential for
external expansion during the inference process. Thus, the general architecture of the
developed fuzzy ES 1 is shown in Figure 4.
More information on the proposed architecture can be found in the authors’ work
on a detailed review of the methodology for creating a fuzzy expert system application
suitable for work in the IT field combined with the Stage-Gate framework [33].

4. Application of fuzzy expert systems in IT project management

On the example of one of the expert consultations, we will demonstrate the work
of an expert system based on fuzzy logic using a combined model of the semantic
network SNePS (Semantic Network Processing System) and fuzzy inference rules.
Consider the following problem statement for an IT company. The model situation
for projects is the availability of five teams (CycleDuo, Templater, Avion, Howl, and
Converge) to develop existing and potential projects. A new team (Emerald) was also
hired during the year, increasing the total number of available teams to six. The
current market situation is to select five projects (Genesis, Crowding, Firantis, Explo-
ration, and Hymera) from two customers (Mazzle, Global State). Also, additional
information is provided on costs not directly related to current projects, namely the
costs inherent in the maintenance of the office, administrative staff, and the cost of
various advertising campaigns. Given the available data, the IT company faces the
task of finding the best way to maximize annual profits.
According to the above problem statement, we will demonstrate the results of
using a fuzzy expert system based on modeled data and compare those results with

1
Source code of developed fuzzy expert system could be found at https://github.com/frightempire/
FuzzyExpert.

9
Project Management - New Trends and Applications

Name Description

Net revenue The difference between the company’s profit and all costs. From an accounting point
of view, net profit is the difference between gross revenue and costs associated with
managing the firm.

Total revenue The total sales of the company and other sources of profit. It is important to note
that total revenue differs from net revenue because it does not take into account
expenses.

Gross revenue Income at invoice value received for goods and services over a period of time.

Cost of goods sold Value of goods or services sold during a certain period.
(COGS)

Operating expenses Amount paid for the maintenance of assets or business expenses, excluding
depreciation.

Table 1.
Net revenue calculation indicators.

real historical data. To prove the success of the proposed method, it is necessary to
find a way to measure its performance. In this example, it is advisable to consider the
criterion of net profit as a measure of the effectiveness of the task of maximizing the
company’s profits [34]. Next, consider in more detail the process of calculating net
revenue.
First, consider the main indicators involved in the calculation of net revenue.
Table 1 provides a general description of these indicators [35].
Let us start with the calculation of gross revenue:

Gross revenue ¼ total revenue COGS, (3)

where the value of total revenue can be obtained by summing the planned annual
profit from projects in development, and the cost of goods sold in our case is the cost
of services, that is, the total cost of compensation for teams working on projects in
development.
Given the value of gross revenue, it is possible to calculate the value of net revenue:

Net revenue ¼ gross revenue operating expenses, (4)

where operating expenses in our case are the costs of advertising campaigns and
assets in the form of office space and administrative staff.
Thus, based on the indicators for which data are available, we can formulate the
following method of calculating net revenue:

Net revenue ¼ total revenue COGS operating expenses: (5)

4.1 Knowledge base creation process

To fill the knowledge base we used a combination of project managers’ expertise

and leading research results in the field of software project management (described in
Section 2). Combining research data with information provided by management
specialists from approximately 30 Ukraine IT companies, which describes the general
structure of business processes during the management of the project portfolio within
10
Application of Fuzzy Expert Systems in IT Project Management
DOI: http://dx.doi.org/10.5772/intechopen.102439

the IT company, a knowledge base was formed. Knowledge base corresponds to the
IF-THEN structure and is aimed at solving a specific task of maximizing the profits of
a typical IT company.
Templates for creating a knowledge base are as follows:

1. IF (Risk of cooperation with the company X is N1 AND Project Y profit from the
company X is N2) THEN ({It makes {no} sense to consider the project Y from the
company X)

2. IF (Company X is the current customer AND It makes sense to consider the project Y
from the company X AND Deviations in company X estimates is N1 AND Project Y
profit type from the company X is N2) THEN (Project Y priority is N3)

3. IF (It makes no sense to consider the project Y from the company X) THEN (Project Y
has no priority)

4. IF (Team X will soon complete the project AND Team X compensation is N1 AND
Risk of interaction with team X is N2) THEN (Team X priority is N3)

5. IF (Team X does not complete the project soon) THEN (Team X has no priority)

6. IF (Team X is without match AND Project Y is without match AND Team X priority
is N1 AND Project Y priority is N2) THEN (Team X has a match AND Project Y has
a match AND Team X corresponds to project Y)

7. IF (Team X has no priority OR Project Y has no priority) THEN (Unable to match

team X to project Y)

8. IF (Team X is without match AND Team X will soon complete the project) THEN
(Team X must be disbanded)

9. IF (Project X is without match) THEN (Need to look for a new team)

10.IF ([Risk of interaction with team X is N]xM OR [...]xM) THEN ({No} risk of
remote work)

11. IF ({No} risk of remote work AND Office expenses is N1) THEN (It makes {no} sense
for remote work transfer)

12. IF (Company size is N1 AND The size of the administrative staff is N2) THEN
(Need to {reduce, no action, increase} administrative staff)

13. IF (Type X advertising campaign expenses is N1 AND The benefits of an type Х

advertising campaign is N2) THEN (Need to {decrease, no action, increase} a type X
advertising campaign)

4.2 Fuzzy expert system implementation results

As a usage result of a fuzzy expert system, the following recommendations were

received from the system:
11
Project Management - New Trends and Applications

• To carry out the Genesis project, it is recommended to select the Howl team.

• To carry out the Hymera project, it is recommended to select the Converge team.

• To carry out the Crowding project, it is recommended to hire a new Emerald

team.

• To carry out the Firantis project, it is recommended to select the Avion team.

• The Exploration project is not recommended to be taken into development.

• It is recommended to abandon the office space.

• It is recommended to expand the budget for an advertising campaign on social

networks.

• It is recommended to leave the budget for the advertising campaign in

universities unchanged.

• It is recommended to abandon the advertising campaign through conferences.

• It is recommended to abandon outdoor advertising.

4.3 Modeled historical data

The available historical data were provided by the HYS Enterprise B.V.2 IT
company and is based on an annual breakdown of data close to real data, namely:

• planned annual revenue from potential projects;

• monthly compensation of current teams and administrative staff;

• annual expenses for maintaining an active office;

• monthly expenses to support active advertising campaigns.

Approximate available data on planned annual revenues from projects can be

found in Table 2. Monthly compensation of teams, annual expenses of supporting
assets in the form of office and administrative staff, as well as monthly expenses of
active advertising campaigns are provided in Table 3.
The value of the annual net revenue was provided already calculated, but for
visualization, we will perform this calculation again. This will help to make a similar
calculation in the case of expert system usage. Before the calculation, we briefly
describe the historically made decisions based on the data described in Tables 1 and 2:

2
All data are not real, but close to real. Any similarity to real data is a coincidence. HYS Enterprise B.V. is
not responsible for the correctness or sharing of the methods proposed in this work, as well as for the
quality of the results obtained on their basis. If any questions about described methods, test data, or results
occur, it is recommend to contact the authors.

12
Application of Fuzzy Expert Systems in IT Project Management
DOI: http://dx.doi.org/10.5772/intechopen.102439

Mazzle

Genesis 350,000$

Crowding 250,000$

Global state

Firantis 285,000$

Exploration 90,000$

Hymera 320,000$

Table 2.
Planned annual revenues from potential projects.

Team compensations (monthly)

CycleDuo 7000$

Templater 4000$

Avion 12,000$

Howl 8000$

Converge 3000$

Emerald 5000$

Asset support expenses (monthly)

Administrative staff 15,000$

Office space 18,000$

Advertising campaign support expenses (annual)

Social networks 12,000$

Conferences 90,000$

Universities 50,000$

Outdoor advertising 15,000$

Table 3.
Expenses by different categories.

• To carry out the Genesis project the Howl team was selected.

• To carry out the Hymera project the Converge team was selected.

• To carry out the Crowding project the Avion team was selected.

• To carry out the Firantis project a new Emerald team was hired.

• The Exploration project was not taken into development.

• It was decided not to abandon the office space.

• The size of administrative staff has been reduced.

• The budget for the advertising campaign on social networks was expanded.

13
Project Management - New Trends and Applications

• The budget for the advertising campaign at universities was expanded.

• The conference budget was left unchanged.

• It was decided to abandon outdoor advertising.

First, we calculate the planned total revenue based on the planned annual revenues
from the projects. Already existing projects are also taken into account in the calcula-
tions regardless of their completion date. Another interesting point is the failure of the
Crowding project by the Avion team due to the combination of high risk of interaction
with the customer Mazzle and the Avion team. Thus, the Crowding project is not
taken into account in the calculations:

Total revenue ¼ 12  1000  ð150 þ 90 þ 50 þ 45 þ 35Þ þ 12  1000  ð350 þ 285 þ 320Þ

¼ 1, 640, 000$:
(6)

Now we calculate the annual cost of goods sold, which in this example is the sum of
the compensation of the teams involved in the development. Teams involved in the
development of current projects that have not yet been completed are also taken into
account in this calculation:

COGS ¼ 12  1000  ð7 þ 4 þ 12 þ 8 þ 3 þ 5Þ ¼ 468, 000$: (7)

It remains to calculate the operating expenses, which consist of advertising cam-

paigns, office support, and administrative staff expenses:

Operating expenses ¼ 12  1000  ð18 þ 15Þ þ 1000  ð12 þ 50 þ 90 þ 15Þ

¼ 563, 000$: (8)

All indicators necessary for calculation of net revenue are prepared:

Net revenue ¼ 1, 640, 000$ 468, 000$ 563, 000$ ¼ 609, 000$: (9)

5. Results

Here we will consider in what aspects the historical solutions coincide and differ
with solutions proposed by the expert system and will demonstrate the effects of those
differences. Let us briefly summarize the differences in decision making:

• To complete the Crowding project, it is proposed to hire a new Emerald team as

opposed to the Avion team.

• To implement the Firantis project, it is proposed to choose the Avion team as

opposed to a new Emerald team.

• It is proposed to abandon the office.

• It is proposed to expand the budget for the advertising campaign on social

networks, in contrast to the historical data, which left this budget unchanged.
14
Application of Fuzzy Expert Systems in IT Project Management
DOI: http://dx.doi.org/10.5772/intechopen.102439

• It is proposed to leave the budget for the advertising campaign in universities

unchanged, in contrast to the historical data that increased this budget.

• It is proposed to abandon the advertising campaign in the form of conferences, in

contrast to the historical data, which left this budget unchanged.

Other expert system decisions are similar to decisions from historical data. The first
major difference is in finding a team for the Crowding project. The expert system
analyzed the risk of cooperation with the customer and existing teams and concluded
that due to a combination of high risks it is less risky to hire a new team to implement the
project than to appoint one of the existing ones. Thus, the probability of successfully
completing the project and avoiding the situation demonstrated in the historical data on
the failure of the Crowding project due to incompatibility with the Avion team increases.
Given this information, the Avion team was tasked with working on the Firantis project.
The next difference is a proposal to abandon the active office. This decision was
made after analyzing the risks of working with current teams and obtaining a low
overall risk. The usefulness of current advertising campaigns is analyzed. The budget
for the social media campaign has been expanded, which on the one hand increases
expenses, but due to high utility and small investments creates the most favorable
environment for finding new teams and customers. This in the long run leads to a
more rapid expansion of the company and increases the likelihood of finding a cus-
tomer, which will increase the value of total revenue. Due to the average level of
utility and costs, it was decided to leave the budget of university advertising cam-
paigns unchanged. After analyzing the low level of utility and high costs of the
campaign through conferences, the expert system made an unequivocal decision to
abandon this type of campaign.
Next let us make similar calculations of net revenue, taking into account the
implementation of the expert system recommendations. But it should be borne in
mind that since the data are historical, the implementation of the expert system is
modeled. Based on these calculations, we can observe the impact of these decisions on
the value of total profit and operating expenses.
We calculate total revenue based on planned annual revenues from existing and
potential projects. In contrast to real historical data, the Crowding project corresponds
to a team with a lower risk of cooperation, which suggests a higher probability of
successful completion of the project. Thus, in this calculation, the Crowding project is
taken into account:

Total revenue ¼ 12  1000  ð150 þ 90 þ 50 þ 45 þ 35Þ þ 12  1000  ð350 þ 250 þ 285 þ 320Þ
¼ 1, 890, 000$:
(10)

The calculation of the compensation amount of the teams involved in the devel-
opment does not differ from the calculation of historical data. After all, the same
teams are involved in the development. The amount of compensation is:

COGS ¼ 468, 000$: (11)

Thus, we immediately proceed to the calculation of operating expenses. The dif-

ference from historical data is the proposal to abandon the active office, as well as the
advertising campaign through conferences. At the same time, do not change the

15
Project Management - New Trends and Applications

budget for university advertising campaigns. We are reducing the budget for univer-
sity advertising campaigns compared to historical data from $50,000 to $30,000,
which is the size of this budget before expansion. We are expanding the budget for an
advertising campaign on social networks in approximately the same proportions from
$12,000 to $20,000. Refusing an advertising campaign through conferences and office
support, we generally have:

Operating expenses ¼ 12  1000  15 þ 1000  ð20 þ 30Þ ¼ 230, 000$: (12)

All indicators necessary for calculation of net revenue are prepared:

Net revenue ¼ 1, 890, 000$ 468, 000$ 230, 000$ ¼ 1, 192, 000$: (13)

Comparing historical data and implementation results, we can see that the total
profit increased by $250,000. In turn, operating expenses decreased by $333,000.
Thus the total increase in annual net revenue is $583,000.

6. Conclusions

The role of the main subjects in the industry—IT companies—and the level of
management of their business processes are growing. Effective management will help
to achieve the strategic goals of companies and strengthen their financial stability. The
analysis of the instrumental base for support of management decision-making in the
conditions of uncertainty and risk brings to the fore the use of expert systems. At the
same time, fuzzy expert systems built using methods and models of fuzzy logic seem
to be the most effective.
Historical review of research and applications of this mathematical apparatus has
shown only some examples of its use in the field of IT—a “bottleneck” that must be
overcome because the feasibility of implementing intelligent technologies is
confirmed by many factors. Among the main ones—the ability to present available
information in linguistic form, smoothing insufficient or missing information, the
institutional memory of the ES with tools to supplement and modify it, the presence
of built-in mechanisms (for various architectures and algorithms) decision-making,
metacomponent to explain expert advice, a library of precedents for adjusting
the adoption of previous management decisions with the involvement of expert
data, etc.
Given the need and feasibility of implementing the apparatus of fuzzy ES in IT
project management, a fuzzy ES application was developed. The main difference
between the developed application and the existing one is the proposal of the archi-
tectural model of the ES with a modified mechanism of fuzzy inference. This signifi-
cantly speeds up the process of fuzzy inference and reduces the duration of expert
consultations. The effectiveness of the proposed fuzzy ES application is demonstrated
by the example of a modeled situation of maximizing the revenue of an IT company in
specific circumstances—the business process environment associated with the imple-
mentation of current projects.
Currently, the experimental operation of the developed expert system application
is carried out on the basis of a number of IT companies. Topics of expert advice on IT
business process management include a wide range of tasks, such as search and
monitoring of projects (selection of customers, teams, and other resources), forming a
16
Application of Fuzzy Expert Systems in IT Project Management
DOI: http://dx.doi.org/10.5772/intechopen.102439

balanced portfolio of the company, assessing the status of projects, ensuring strategic
goals and improving (in particular, financial) indicators of its activity, etc.
The experience gained based on the sector of outsourced IT companies allowed us
to propose a methodology for embedding the ES application in the overall manage-
ment loop within the Stage-Gate framework, which increases the efficiency of the
system.
Further improvement of the application is carried out in the following areas:

• Expansion of knowledge and databases through definitions that meet the

specifics of the IT industry functioning.

• Improving the accuracy and adequacy of the implemented fuzzy inference due
to:

◦ Combining fuzzy and crisp calculations;

◦ Integration with the metacomponent of historical analysis;

◦ Increasing the flexibility of the fuzzy model for knowledge base creation.

• Creation of a more developed metacomponent to explain in detail the expert

consultations results.

• Development of a user-friendly interface for end-users, taking into account the

wishes of IT professionals.

In essence, expert systems belong to the reusable apparatus. Their effectiveness as

a management tool increases with the enrichment of knowledge and databases
through the introduction of new experiences. The rapid development of the IT indus-
try, specification, and standardization of software development processes at the global
level provides a fundamental basis for the exchange, reproduction, and implementa-
tion of intelligent technologies with elements of fuzzy logic.

Acknowledgements

We would like to show gratitude to HYS Enterprise B.V.™ for providing modeled
historical data for the purposes of this research and validation of the proposed fuzzy
ES and its practical relevance.

17
Project Management - New Trends and Applications

Author details

Oleksii Dudnyk* and Zoia Sokolovska

Department of Economic Cybernetics and Information Technologies, Odessa National
Polytechnic University, Odessa, Ukraine

*Address all correspondence to: [email protected]

© 2022 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms
of the Creative Commons Attribution License (http://creativecommons.org/licenses/
by/3.0), which permits unrestricted use, distribution, and reproduction in any medium,
provided the original work is properly cited.
18
Application of Fuzzy Expert Systems in IT Project Management
DOI: http://dx.doi.org/10.5772/intechopen.102439

References

[1] Digital Economy Report 2019 International Journal of Project

[Internet]. UNCTAD; 2019. Available Management. 2019;37(8):989-1002.
from: https://unctad.org/system/files/ DOI: 10.1016/j.ijproman.2019.09.001
official-document/der2019_en.pdf
[Accessed: December 04, 2021] [8] Lin L, Müller R, Zhu F, Liu H.
Choosing suitable project control modes
[2] Portman H. Review Standish Group— to improve the knowledge integration
CHAOS 2020: Beyond Infinity under different uncertainties.
[Internet]. 2021. Available from: https:// International Journal of Project
hennyportman.wordpress.com/2021/01/ Management. 2019;37(7):896-911.
06/review-standish-group-chaos-2020- DOI: 10.1016/j.ijproman.2019.07.002
beyond-infinity/ [Accessed:
December 04, 2021] [9] Bjorvatn T, Wald A. Project complexity
and team-level absorptive capacity as
[3] The Standish Group. CHAOS Report drivers of project management
2015 [Internet]. 2015. Available from: performance. International Journal of
https://www.standishgroup.com/sample_ Project Management. 2018;36(6):876-888.
research_files/CHAOSReport2015-Final. DOI: 10.1016/j.ijproman.2018.05.003
pdf [Accessed: December 04, 2021]
[10] Keil M, Rai A, Cheney Mann JE,
[4] Tam C, Moura EJ da C, Oliveira T, Zhang GP. Why software projects
Varajão J. The factors influencing the escalate: The importance of project
success of on-going agile software management constructs. IEEE
development projects. International Transactions on Engineering
Journal of Project Management. 2020; Management. 2003;50(3):251. DOI:
38(3):165-176. DOI: 10.1016/j.ijproman. 10.1109/TEM.2003.817312
2020.02.001
[11] Tavares BG, da Silva CES, de
[5] Fink L, Pinchovski B. It is about time: Souza AD. Risk management analysis in
Bias and its mitigation in time-saving Scrum software projects. International
decisions in software development Transactions in Operational Research.
projects. International Journal of Project 2019;26(5):1. DOI: 10.1111/itor.12401
Management. 2020;38(2):99-111.
DOI: 10.1016/j.ijproman.2020.01.001 [12] Rosenau MD, Githens GD. Successful
Project Management: A Step-By-Step
[6] Hoffmann D, Ahlemann F, Reining S. Approach with Practical Examples.
Reconciling alignment, efficiency, and 4th ed. United States: Wiley; 2005
agility in IT project portfolio management:
Recommendations based on a revelatory [13] Mirza MN, Pourzolfaghar Z,
case study. International Journal of Project Shahnazari M. Significance of scope in
Management. 2020;38(2):124-136. DOI: project success. Procedia Technology.
10.1016/j.ijproman.2020.01.004 2013;9:722-729. DOI: 10.1016/j.protcy.
2013.12.080
[7] Einhorn F, Marnewick C, Meredith J.
Achieving strategic benefits from [14] Bloch M, Blumberg S, Laartz J.
business IT projects: The critical Delivering large-scale IT projects on
importance of using the business case time, on budget, and on value. McKinsey
across the entire project lifetime. Quarterly. 2012;27:2-7
19
Project Management - New Trends and Applications

[15] Charette RN. Software Engineering 29(2):300-309. DOI: 10.1016/j.eswa.

Risk Analysis and Management. USA: 2005.04.003
McGraw-Hill, Inc.; 1989
[24] Zarandi MHF, Asl AAS, Sotudian S,
[16] Pfleeger SL, Atlee JM. Software Castillo O. A state of the art review of
Engineering: Theory and Practice. intelligent scheduling. Artificial
4th ed. USA: Pearson; 2010 Intelligence Review. 2020;53(1):501-593.
DOI: 10.1007/s10462-018-9667-6
[17] Arnuphaptrairong T. Top ten lists of
software project risks: Evidence from the [25] Zadeh LA. The role of fuzzy logic in
literature survey. In: Proceedings of the the management of uncertainty in expert
International MultiConference of systems. Fuzzy Sets and Systems. 1983;
Engineers and Computer Scientists. 11(1-3):199-227. DOI: 10.1016/S0165-
Hong Kong: IAENG (International 0114(83)80081-5
Association of Engineers); 2011. pp. 1-6
[26] Dufner D, Kwon O, Doty A.
[18] Alba E, Francisco Chicano J. Improving software development project
Software project management with GAs. team performance: A Web-based expert
Information Sciences. 2007;177(11): support system for project control. In:
2380. DOI: 10.1016/j.ins.2006.12.020 Proceedings of the Hawaii International
Conference on System Sciences.
[19] Uzzafer M. A simulation model for Manhattan, New York, U.S.: IEEE Comp
strategic management process of Soc; 1999. p. 6. DOI: 10.1109/
software projects. Journal of Systems and HICSS.1999.772622
Software. 2013;86(1):21. DOI: 10.1016/j.
jss.2012.06.042 [27] Dufner D, Kwon O, Hadidi R.
Development and implementation of a
[20] Newell A, Shaw JC, Simon HA. WEB-enabled CyberCollaboratory for
Report on a general problem solving asynchronous teams (Web-CCAT). In:
program. In: Proceedings of the Proceedings of the Thirty-First Hawaii
International Conference on Information International Conference on System
Processing. Santa Monica, California, Sciences. Kohala Coast, HI, USA: IEEE;
U.S.: RAND Corporation; 1959. 1998. pp. 330-339. DOI: 10.1109/HICSS.
pp. 256-264 1998.653116

[21] Feigenbaum E, Buchanan B, [28] Truicǎ CO, Barnoschi A. Innovating

Lederberg J. On generality and problem HR using an expert system for recruiting
solving: A case study using the IT specialists—ESRIT. Journal of
DENDRAL program. Machine Software & Systems Development. 2015;
Intelligence. 1970;6:1 2015:1-11. DOI: 10.5171/2015.762987

[22] McNeilly M, Gessner S. Business [29] Rodríguez A, Ortega F,

Insights™: An expert system for Concepción R. A method for the
strategic analysis. Planning Review. evaluation of risk in IT projects. Expert
1993;21(2):32-33. DOI: 10.1108/ Systems with Applications. 2016;45:
eb054410 273-285. DOI: 10.1016/j.eswa.2015.
09.056
[23] Rao MP, Miller DM, Lin B. PET: An
expert system for productivity analysis. [30] Shapiro SC, Rapaport WJ. The
Expert Systems with Applications. 2005; SNePS family. Computers and
20
Application of Fuzzy Expert Systems in IT Project Management
DOI: http://dx.doi.org/10.5772/intechopen.102439

Mathematics with Applications. 1992;23

(2-5):243-275. DOI: 10.1016/0898-1221
(92)90143-6

[31] Schlegel DR, Shapiro SC. Concurrent

reasoning with inference graphs. In:
Lecture Notes in Computer Science
(Including Subseries Lecture Notes in
Artificial Intelligence and Lecture Notes
in Bioinformatics). New York, U.S.:
Springer Verlag; 2014. pp. 138-164. DOI:
10.1007/978-3-319-04534-4_10

[32] Choi J, Shapiro SC. Efficient

implementation of non-standard
connectives and quantifiers in deductive
reasoning systems. In: Proceedings of the
Twenty-Fifth Hawaii International
Conference on System Sciences. Vol. 1.
Manhattan, New York, U.S.: Institute of
Electrical and Electronics Engineers
(IEEE); 1992. pp. 381-390. DOI: 10.1109/
HICSS.1992.183186

[33] Sokolovska Z, Dudnyk O. Devising a

technology for managing outsourcing
IT-projects with the application of fuzzy
logic. Eastern-European Journal of
Enterprise Technologies. 2021;2
(3(110)):52. DOI: 10.15587/
1729-4061.2021.224529

[34] Net revenue [Internet]. Farlex

Financial Dictionary. Available
from: https://financial-dictionary.
thefreedictionary.com/Net+revenue
[Accessed: November 07, 2021]

[35] Farlex Financial Dictionary

[Internet]. Available from https://
financial-dictionary.thefreedictionary.
com [Accessed: November 07, 2021]

21