Automation of Service Desk: Knowledge Management Perspective

Michal Dost

and Jan Skrbek

Department of Informatics, Faculty of Economics, Technical University of Liberec, Czech Republic

Keywords:

Service Desk, Automation, Knowledge Management.

Abstract:

In the current times, the need for quality and effective service support technologies is high. We can achieve

good IT service operations that support good knowledge management practices by employing automation

techniques and methods to the Service Desk systems. In our position paper, we look at the literature regarding

Service Desk automation in relation to knowledge management. We then propose a theoretical model of

Automatized Service Desk Systems that uses several techniques to help and automatize the process of user

request resolution. Our proposed system employs text mining techniques, a virtual agent, an expert system, a

customer intent detector, and a ticketing system.

1 INTRODUCTION

Nowadays, more than ever, the emphasis on mak-

ing access to services more accessible and optimized

is gaining importance. With Service Desk, it is not

different. Employees working from home require a

quality Service Desk, and therefore the companies are

looking at ways to optimize and level up their IT ser-

vice operations. One way to achieve this is to auto-

mate some processes involved in the Service Desk op-

erations.

This position paper proposes a theoretical model

of an automated service desk system, which includes

several technologies that enable its effective automa-

tion related to practical knowledge management.

For our literature review, we set out several search

parameters: research articles and other academic liter-

ature published in the last 15 years (however, we also

used some sources published before the year 2006).

We searched for the topics related to automation of

the Service Desk. Based on our literature research,

we then prepared and described the theoretical model

of the Automated Service Desk System.

Our position paper is structured as follows: Sec-

tion 2 describes main topic areas of automation in

Service Desk and Knowledge Management. We also

look at some implementations of expert systems in the

Service Desk environment as it is a way to automatize

some tasks, and it is also a knowledge-based system.

In Section 3 we then describe the proposed automated

system.

https://orcid.org/0000-0001-6398-624X

2 LITERATURE REVIEW

In this section, we describe the current literature re-

garding the topic of Service Desk automation.

2.1 Knowledge Representation

When working with the concept of automated ser-

vice desk process, one of the essential topics from

the knowledge management perspective is the way,

which will be the knowledge represented. This brings

out the question: What will be the optimal knowl-

edge representation for humans involved in the pro-

cess, and what would be optimal for the system?

According to (Czarnecki and Sitek, 2013) there

are two types of symbolic knowledge representation:

(1) procedural representation, through which a set of

procedures is identiﬁed that describe the operations

behind the knowledge; and (2) declarative represen-

tation, which uses sets of statements, facts, and rules

from the particular problem domain. In the Service

Desk environment, there are both types of knowledge

and our proposed system should be able to work with

both. The procedural representation may describe

speciﬁc actions in order to resolve speciﬁc user re-

quests.

The academic literature is working with knowl-

edge representation for quite some time. The topic

is strongly connected to expert systems, where the

representation of knowledge is essential. Expert sys-

tems are a subtopic of Artiﬁcial Intelligence, which

as a science was founded in 1956 in Dartmouth and

204

Dostál, M. and Skrbek, J.

Automation of Service Desk: Knowledge Management Perspective.

DOI: 10.5220/0010693100003064

In Proceedings of the 13th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2021) - Volume 3: KMIS, pages 204-210

ISBN: 978-989-758-533-3; ISSN: 2184-3228

where began the development of the ﬁrst expert sys-

tem called Logic Theorist (Watson and Mann, 1988).

(Jakus et al., 2013) states that the most widespread

medium of communication is in natural language;

however, it is not appropriate for the needs of auto-

mated or intelligent systems. There are several knowl-

edge representation formalisms that are more con-

venient for use with computer systems. Those are

semantic networks, conceptual graphs, fuzzy logic,

frames or description logics (Jakus et al., 2013). An-

other types of representation of knowledge are classi-

cal logic (propositional, ﬁrst and second-order logic),

constraint logic programming, ontologies, Bayesian

networks (Porter et al., 2008).

When looking at knowledge representation in IT

Service Management, in the literature, we can ﬁnd

some examples. (Paramesh et al., 2018) wrote about

a classiﬁer for Service Desk tickets that are, by their

nature, unstructured. For their classiﬁer, they used

methods of machine learning. The unstructured data

were pre-processed by cleaning the raw data from un-

wanted entries. Then they constructed a feature vec-

tor for each entry, which acts as a representation of

the Service Desk ticket data. For building the clas-

siﬁcation models, they used Bootstrap Aggregation,

Boosting, and Voting Ensemble for the predictions

from different models to be combined.

In their following work, (Paramesh and Shreed-

hara, 2019) further described the automated system

for Service Desk. The representation of knowledge

by feature vectors is achieved by applying the TF-

IDF (Term Frequency - Inverse Document Frequency)

term weighting scheme. The features are then se-

lected by χ

test to reduce dimensionality.

Authors (Czarnecki and Sitek, 2013) dealt with

the comparison between description logic based on-

tologies and rule based approach of knowledge rep-

resentation with the connection to IT Service Man-

agement. This type of representation is classiﬁed as a

declarative representation of knowledge.

In the older literature, we can also ﬁnd some au-

tomation applications in Service Desk, and what is

more important, ways of representing the knowledge.

For example, (Greer et al., 1998) described an intel-

ligent help desk system that worked with a type of

concept map that was divided into two layers.

Next, (Chan et al., 2000) described intelligent

case-based system for help desk operations. The sys-

tem was built on case-based reasoning technology, so

the knowledge used by the system was represented in

the form of cases. Speciﬁcally, the cases consisted

of special objects (folders, cases, questions, and ac-

tions).

In our proposed system, there are multiple rep-

resentations of knowledge with which the system

works. During the text mining phase, the input is

transformed into a set of keywords in the form of an

object. The knowledge base then contains knowledge

in the form of rules as the expert system is rule-based.

2.2 Knowledge Discovery

In this subsection of our position paper, we talk about

methods of automation that can be applied to knowl-

edge discovery activities. We can describe knowl-

edge discovery as an activity or process that aims to

the production of knowledge by discovering or deriv-

ing from existing information (Soundararajan et al.,

2005). The process of knowledge discovery consists

of: selection of target data, data cleaning, preprocess-

ing, reduction, projection of the data, preparation for

data mining by choosing the correct function and al-

gorithm; primary data mining process and after that,

an interpretation and usage of the discovered knowl-

edge (Soundararajan et al., 2005).

One of the most popular techniques of knowledge

discovery today is data and text mining. Text min-

ing, also referred to as Knowledge Discovery in text,

is deﬁned as ”... the art of data mining from text data

collection” (Cai and Sun, 2009). Text mining aims to

discover information and knowledge in unstructured

or semi-structured forms of text data. There are sev-

eral techniques used for text mining: (1) information

extraction, (2) information retrieval, (3) categoriza-

tion, (4) clustering, and (5) summarization.

Text mining nowadays has a stable role in the

Service Desk systems that aim to automate the pro-

cess of ticket resolution. One example is the system

described by (Al-Hawari and Barham, 2021), which

uses a machine learning model allowing automatic

ticket classiﬁcation.

The exciting use of text mining is its applica-

tion on a conversation at a call center, described by

(Takeuchi and Yamaguchi, 2014). The telephone con-

versations are transcribed to their textual form by an

automatic speech recognition program—this serves

as a prerequisite for a call summarization, which is

performed later. According to (Takeuchi and Yam-

aguchi, 2014) it is essential to identify expressions in

the speech transcription that are important for the con-

versation to be correctly classiﬁed and summarized.

Their system has two main components. The ﬁrst one

is used to extract words, tag them and assign seman-

tic categories. The extracted words are also compared

with an expert knowledge dictionary for the check of

their importance.

Automation of Service Desk: Knowledge Management Perspective

205

(Agarwal et al., 2017) described a system for au-

tomatic problem extraction and the analysis of text in

tickets created in the IT Service Desk. Those tickets

contain unstructured data, and a given problem de-

scription can be relatively ”noisy.” There is always a

high possibility that such data will contain misspelled

words or words in the form of abbreviations. They

used a very interesting approach of extracting the log-

ical structure of the IT ticket by categorizing the word

into two groups: category dependent and category in-

dependent. The category dependent contains context-

deﬁning words, generic and speciﬁc words, or pat-

terns. The category independent words are catego-

rized as domain invariant words or domain invariant

patterns. By doing this, they can ﬁlter out irrelevant

words that may change to correct understanding and

classiﬁcation of the text. They used a support vector

machine with RBF Kernel (Radial Basis Function) for

the classiﬁcation engine.

IT Service Desk tickets can also be used for dis-

covering knowledge about the customers, e.g., their

satisfaction with the product. The approach of (Eck-

stein et al., 2016) not only classiﬁes if the customer

is satisﬁed but also tries to discover what their needs

are. Their system uses a text mining process called

”bag of words” and techniques such as tokenization,

stemming, and the removal of stop words. For the

training of the classiﬁcation models, they used deci-

sions trees, support vector machines, kNN (k-nearest

neighbors), and Naive Bayes.

2.3 Methods of Automation

In this subsection, we look at methods of Service

Desk automation described in the literature.

(Mani et al., 2017) described a system for question

answering that is used in the IT support environment.

It has interesting capabilities, such as the ability to

process unstructured documents such as web pages,

PDF ﬁles, and audio and video ﬁles. They also de-

scribe an ability of ”effective addition of human-in-

the-loop,” a way to delegate a speciﬁc task that the

system cannot complete to a speciﬁc person. This

relates to Expert Locator Systems, which store em-

ployee information and knowledge about company

problem domains. The described system can also

transcribe audio ﬁles, which could also be helpful in

the call center environment.

The next topic mentioned in the literature regard-

ing automation of processes in the IT Service Desk is

virtual agents. (Lacity et al., 2017) describes a virtual

agent employed in the systems of the Bank of Swe-

den. This system can change the password on behalf

of the user’s request, and it can unlock their locked

accounts or provide access to certain documents.

The use of an intelligent agent in Service Desk

processes is also described by (Koehler et al., 2018).

They used machine learning techniques to ﬁlter out

the non-relevant parts of the request. This way, they

isolate vital information about the user’s problem.

(Ali, 2018) describes the use of cognitive tech-

nologies as a way of optimizing the IT services, such

as Service Desk. Their cognitive system uses methods

such as semantic analysis of the tickets, NLP (Natural

Language Processing), or extraction of structured and

unstructured data for in-depth analysis and visualiza-

tion.

To the topic Service Desk automation also con-

tribute topics such as user proﬁling, personalization or

stereotypization (Zaslavsky et al., 2007), expert pro-

ﬁling. (Baysal et al., 2009).

2.4 Expert Systems in Service Desk

An expert system is a system able to emulate the

knowledge of an expert in a particular problem do-

main. There are two main categories of expert sys-

tems: diagnostic and planning expert systems. The

ﬁrst ones are used for data interpretation with the aim

of the best correspondence with actual data. The lat-

ter is used for problem-solving, as they try to generate

possible solutions to a given problem. Both of these

types of systems can be used in a Service Desk envi-

ronment.

Fundamental parts of an expert systems are knowl-

edge base and inference mechanism. There is also

some form of explanation module and communica-

tion module (or user interface) in many cases.

An example of the usage of expert systems in

the IT Service Desk is described by (Al-Emran and

Al Chalabi, 2015). They developed a troubleshooter

expert system focused on three problem domains:

printers, HW+SW problems, and problems with inter-

net connection. For building their system, they used

CLIPS, a rule-based building tool that uses a syntax of

LISP programming language. They created more than

70 rules to cover the possible topics of three problem

domains.

(Songsangyos et al., 2012) proposed a prototype

of help desk service that uses an expert system. The

knowledge base is built with production rules, and

therefore it is a rule-based expert system. The pro-

posed system can solve a range of problems with

computer hardware, software, or problems with the

network. It can also identify the root causes of the

problems and provide a step-by-step solving guide.

The system searches the knowledge base for the rules

that contain the ”goal.” If it ﬁnds any, the system re-

KMIS 2021 - 13th International Conference on Knowledge Management and Information Systems

206

turns a list of possible causes.

(Kaushik et al., 2011) proposed an expert system

called Network Expert System (NES), which is aimed

at troubleshooting problems with computer networks.

The proposed system, NES, is classiﬁed as a rule-

based diagnostic system. This type of system is not

implemented in the Service Desk system; however, it

has a great potential to be implemented in one.

(Bello et al., 2018) presented an expert system

called ExperTI, which automatically assists the users

(customers) and the Service Desk operators. The use

of web chat enables this. The knowledge base is built

on ontologies and rules.

3 AUTOMATED SERVICE DESK

SYSTEM

This section of our paper discusses the possible impli-

cations and constraints of implementing our theoreti-

cal automated service desk system. In the following

subsections, we describe each component of the sys-

tem and its connection to knowledge management.

Let us begin with an overall description of the

system. At the core of the automated Service Desk

system, there is a virtual agent, which handles the

whole operation of the system with the supervision of

the system manager, who is a member of the Service

Desk staff. Other actors in this use case are:

• User - the one, who is contacting the Service Desk

Department with their request or question

• Operator - a member of Service Desk staff,

tasked with resolution of the tickets and requests,

that could not be resolved by the system; the oper-

ator is also instance of an expert, who is contacted

when needed

• Problem Domain Owner - a member of Service

Desk staff who administers a certain portion of the

Knowledge Base, according to their domain of ex-

pertise

• Manager - a Service Desk manager, who over-

sees the correct running of all processes in and

out of the system

Three types of interfaces are used for the commu-

nication and administration of the system. The ﬁrst

interface is through a telephony system, in which case

the user calls to Service Desk by phone. At ﬁrst, the

user is greeted by the virtual agent, who handles in-

coming information from the user. The same goes for

the other two interfaces, which are computer-based -

one is a chat interface, which could be facilitated by

some speciﬁc Instant Messaging software or web ap-

plication. A third interface is a Web Form. This last

type of communication is not in real-time as the user

sends the request to Service Desk via Web Form and

is later either contacted by the system or speciﬁc ex-

pert or operator about successful resolution of their

request or is asked some follow up questions that may

streamline the resolution of the user‘s request.

As the reader can see in Figure 1, the information

from the user is ﬁrst going through a text mining mod-

ule, which can ﬁlter out the non-relevant parts of the

text or transcribed speech. The module tries to de-

tect the core of the user‘s request and prepare it for

processing by the virtual agent.

The virtual agent then cooperates with other mod-

ules of the system. One of them is the Customer Intent

Detector, which task is to detect the emotions and in-

tent of the user. This information can be used to alter

the communication style and later be used to retro-

spectively analyze the interaction and create a space

for improvement of the process.

The next module of the system is the Expertise Lo-

cator System, which is employed if the Virtual Agent

cannot resolve the opened ticket. Expertise Loca-

tor System contains knowledge and information about

employees and their expertise. These employees also

may not be from the IT Service Desk department and

can be staff members of any department of the com-

pany or corporation. In some cases, there may also

be contact information to a former staff member, who

was a long-time employee and therefore poses a good

portion of the needed knowledge.

The embedded Expert System is then used as a di-

agnostic tool for investigating problems that belong

to a speciﬁc problem domain, which knowledge is

present in the knowledge base. The expert system

works with a combination of types of knowledge rep-

resentation. The aim is to be compatible with all mod-

ules of our Automated Service Desk System, as many

of these modules work with the Knowledge Base.

The Knowledge Base is curated and administrated

by the Problem Domain Owners. The owners are ex-

perts in a given area of expertise, whose knowledge

is present in the Knowledge Base. The number of

Problem Domain Owners depends on the categories

of problems that the Service Desk System can solve

and help with.

The Knowledge Base is also connected with a

ticketing system that works automatically and is occa-

sionally edited by the operators or experts. The tickets

also act as a knowledge item and contain helpful in-

formation for later analysis or future user request res-

olution. Based on this knowledge, the virtual agent

can be improved and can learn.

Automation of Service Desk: Knowledge Management Perspective

207

Figure 1: Model of theoretical Automated Service Desk System.

Following is a speciﬁc example of the user’s in-

teraction with the Automated Service Desk System.

Suppose that the user needs to contact the Service

Desk in the matter of not being able to log in to a spe-

ciﬁc module of the company’s information system. In

the ﬁrst step, they chose the tool for communication,

for example, a phone. So they call a speciﬁc tele-

phone number mapped by the telephony system to

redirect the call to an Automated Service Desk Sys-

tem and its Virtual Agent module. The interaction is

initiated by the system asking the user to state their

problem or request. In our case, it might look like:

“Hello, this is the Automated Service Desk System.

Please state your request.” The user then responds

with the description of their request: “Hi, I am try-

ing to login to the XYZ module of our information

system. My username is xjohn001. The system re-

turns an alert that I do not have sufﬁcient rights; how-

ever, yesterday it worked.” Through its text mining

capabilities, the system can detect speciﬁc keywords

and phrases. It extracts the following keywords: lo-

gin, XYZ module, our information system, the user-

name is xjohn001, not have sufﬁcient rights. Based

on this keyword, the system now knows that the prob-

lem concerns access rights to a speciﬁc module for the

user xjohn001. Every interaction between the system

and the user is run through Customer Intent Detec-

tor to detect any unwanted emotions or patterns from

the user. The Virtual Agent module then employs the

Expert System, which works the knowledge base and

detects if the system already resolved a similar prob-

lem in the past. Suppose it was, and the expert system

ﬁnds an entry in the knowledge base. The informa-

tion about the resolution of such request states that

the access rights need to be assigned by an authorized

employee. So in the next step, the system activates

Expertise Locator System to choose an appropriate

employee/operator. If there are two or more autho-

rized operators, the system also considers their cur-

rent occupancy by other requests and selects the most

available one. When the suitable operator is chosen,

the system creates a transcription of the interaction

with the user and prepares a summary consisting of

the main keywords. This data is then transferred to

the authorized operator tasked with the request resolu-

tion, and a ticket is created. The call is now delegated

to the operator, who will verify if the user is eligible

to access the requested information system module.

After successful resolution, the ticket is marked as re-

solved, and a corresponding entry in the knowledge

base is created.

3.1 Knowledge Management

Perspective of the System

The proposed theoretical Automated Service Desk

System is, by its nature, an instance of a knowledge-

based system. The core asset of this system is the

knowledge that it creates, stores, and applies. What

needs to be taken into account when such a system is

developed?

KMIS 2021 - 13th International Conference on Knowledge Management and Information Systems

208

One of the ﬁrst essential questions one must an-

swer is what type of knowledge representation will be

used? In our literature review part of this paper, we

mentioned what representation types are used in simi-

lar systems. A classic approach would be to use rules;

however, it would be more beneﬁcial to use some on-

tology or concept maps due to the nature of Service

Desk data. It is essential to consider that the pro-

posed system has multiple modules that work with the

knowledge stored in the Knowledge Base; therefore,

they must be able to effectively and correctly work

with the knowledge.

To ensure the quality of the knowledge stored in

the knowledge base, we propose for the individual

types of the knowledge domains to be ”owned” by a

speciﬁc expert in that ﬁeld, who is also a member of

the staff of the company or corporation. This way, the

expert can supervise and control the knowledge store

and ensure that it is appropriately stored and the in-

formation provided by the Knowledge Base entry is

sufﬁcient for both the system and its users (incl. op-

erators). These tasks should also be covered and en-

couraged by the company’s Knowledge Management

Department or its Service Desk subdivision.

4 CONCLUSIONS

The role of automation in the Service Desk is indis-

pensable and will keep being important in the future.

Thanks to employing automation methods, the pro-

cesses of the Service Desk and related knowledge

management can be optimized and improved. In our

position paper, we proposed a theoretical model of

an Automated Service Desk System that consists of

a number of modules. We would like to build a proto-

type of such a system and perform feasibility tests in

our future research.

REFERENCES

Agarwal, S., Aggarwal, V., Akula, A. R., Dasgupta, G. B.,

and Sridhara, G. (2017). Automatic problem extrac-

tion and analysis from unstructured text in IT tick-

ets. IBM Journal of Research and Development,

61(1):4:41–4:52.

Al-Emran, M. and Al Chalabi, H. (2015). ”Developing an

IT Help Desk Troubleshooter Expert System for diag-

nosing and solving IT Problems”. In Mobility, Intelli-

gent Networks and Smart Societies.

Al-Hawari, F. and Barham, H. (2021). A machine learning

based help desk system for IT service management.

Journal of King Saud University - Computer and In-

formation Sciences, 33(6):702–718.

Ali, A. R. (2018). Cognitive Computing to Optimize IT

Services. In 2018 IEEE 17th International Confer-

ence on Cognitive Informatics & Cognitive Comput-

ing (ICCI*CC), pages 54–60. IEEE.

Baysal, O., Godfrey, M. W., and Cohen, R. (2009). A bug

you like: A framework for automated assignment of

bugs. In 2009 IEEE 17th International Conference on

Program Comprehension, pages 297–298.

Bello, A., Melgar, A., and Pizarro, D. (2018). ExperTI: A

knowledge based system for intelligent service desks

using free text. Advances in Intelligent Systems and

Computing, 721:397–406.

Cai, Y. and Sun, J.-T. (2009). Text Mining. In Liu, L. and

Ozsu, M. T., editors, Encyclopedia of Database Sys-

tems, volume 2009, pages 3061–3065. Springer US,

Boston, MA.

Chan, C. W., Chen, L.-L., and Geng, L. (2000). Knowl-

edge engineering for an intelligent case-based system

for help desk operations. Expert Systems with Appli-

cations, 18(2):125–132.

Czarnecki, A. and Sitek, T. (2013). Ontologies vs. Rules

— Comparison of Methods of Knowledge Represen-

tation Based on the Example of IT Services Manage-

ment. pages 99–109.

Eckstein, L., Kuehl, N., and Satzger, G. (2016). Towards

Extracting Customer Needs from Incident Tickets in

IT Services. In 2016 IEEE 18th Conference on Busi-

ness Informatics (CBI), volume 01, pages 200–207.

Greer, J., McCalla, G., Cooke, J., Collins, J., Kumar, V.,

Bishop, A., and Vassileva, J. (1998). The Intelli-

gent Helpdesk: Supporting Peer-Help in a University

Course. In Goos, G., Hartmanis, J., van Leeuwen,

J., Goettl, B. P., Halff, H. M., Redﬁeld, C. L., and

Shute, V. J., editors, Intelligent Tutoring Systems, vol-

ume 1452, pages 494–503. Springer Berlin Heidel-

berg, Berlin, Heidelberg.

Jakus, G., Milutinovi

c, V., Omerovi

c, S., and Toma

S. (2013). Knowledge Representation, pages 47–62.

Springer New York, New York, NY.

Kaushik, A., Barnela, M., Khanna, S., and Kumar, H.

(2011). A Novel Expert System for PC Network Trou-

bleshooting and Maintenance. International Journal

of Advanced Research in Computer Science, 2(3).

Koehler, J., Fux, E., Herzog, F. A., L

otscher, D., Waelti,

K., Imoberdorf, R., and Budke, D. (2018). To-

wards Intelligent Process Support for Customer Ser-

vice Desks: Extracting Problem Descriptions from

Noisy and Multi-lingual Texts. pages 36–52.

Lacity, M., Willcocks, L., and Craig, A. (2017). Service

Automation: Cognitive Virtual Agents at SEB Bank.

The Outsourcing Unit Working Research Paper Series,

(February):1–29.

Mani, S., Gantayat, N., Aralikatte, R., Gupta, M., Dechu,

S., Sankaran, A., Khare, S., Mitchell, B., Subrama-

nian, H., and Venkatarangan, H. (2017). Hi, how can

I help you?: Automating enterprise IT support help

desks. arXiv:1711.02012 [cs].

Paramesh, S., Ramya, C., and Shreedhara, K. (2018). Clas-

sifying the Unstructured IT Service Desk Tickets Us-

ing Ensemble of Classiﬁers. In 2018 3rd International

Automation of Service Desk: Knowledge Management Perspective

209

Conference on Computational Systems and Informa-

tion Technology for Sustainable Solutions (CSITSS),

pages 221–227.

Paramesh, S. P. and Shreedhara, K. S. (2019). Automated

IT Service Desk Systems Using Machine Learning

Techniques. In Nagabhushan, P., Guru, D. S., Shekar,

B. H., and Kumar, Y. H. S., editors, Data Analytics

and Learning, volume 43, pages 331–346. Springer

Singapore, Singapore.

Porter, B., Lifschitz, V., and Van Harmelen, F., editors

(2008). Handbook of Knowledge Representation.

Foundations of Artiﬁcial Intelligence. Elsevier, Am-

sterdam ; Boston, 1st ed edition.

Songsangyos, P., Niyomkha, W., and Tumthong, S. (2012).

HELPDESK EXPERT. International Journal of Arts

& Sciences, 5(4):137–143.

Soundararajan, E., Joseph, J., Jayakumar, C., and M, S.

(2005). Knowledge Discovery Tools and Techniques.

Takeuchi, H. and Yamaguchi, T. (2014). Text Mining of

Business-Oriented Conversations at a Call Center. In

Yada, K., editor, Data Mining for Service, volume 3,

pages 111–129. Springer Berlin Heidelberg, Berlin,

Heidelberg.

Watson, H. J. and Mann, R. I. (1988). Expert Systems: Past,

Present, and Future. Journal of Information Systems

Management, 5(4):39–46.

Zaslavsky, A., Bartolini, C., Boulmakoul, A., Alahakoon,

O., Loke, S. W., and Burstein, F. (2007). Enhancing

the IT service desk function through unobtrusive user

proﬁling, personalization and stereotyping. In 14th

Workshop of the HP Software Univesity Association,

Garching/Munich.

KMIS 2021 - 13th International Conference on Knowledge Management and Information Systems

210