A Conception Of Multiagent Management System

Of Dispersed Market Information – E-negotiations Area

Leszek Kiełtyka

, Rafał Niedbał

Department of Management Information Systems, Management Faculty

Częstochowa University of Technology

Al. Armii Krajowej 19B, Poland

Abstract. The conception of m

ultiagent system (MAS) as a tool aiding dis-

persed market information management in e-negotiations area was proposed in

this article. The results of the conducted surveys concerning among other things

identification of the application areas of intelligent software agents in the en-

terprises are also presented here. Attention was paid to the role of business ne-

gotiation in market information acquisition. Software environment Agent-

Builder enabling elaboration of the simulating model of the proposed concep-

tion of the system was also described in the present article.

1 Introduction

The electronic markets develop mainly due to technology and computer technology

advances. Particularly important here are: electronic data interchange (EDI), the Inter-

net – the global network, and especially World Wide Web network service and elec-

tronic mail [5], [8].

Market information is a set of all data, information and news, also informal infor-

ation and news acquired through direct contacts, which is essential to prepare mar-

ket strategy and market operations, including operations on international markets.

Ability to compete effectively and maintain own position on the market may be se-

cured for the enterprise by an efficient system of market information. Thus, there are

initiatives undertaken that aim to develop new access technology to dispersed and

heterogeneous market information. Despite the fact that contemporary IT technology

(Information Technology) offers great opportunity of quick and easy access to vital

information from the point of view of business operations [11], finding required at

a given moment information on the Internet is still troublesome to some managers and

is a complicated and time consuming process.

One of the interesting conceptions is the application of intelligent software agents

operat

ing on behalf of the users [2], [3], [10]. Their task is among other things to

search in many sources for information that meets set in advance criteria, and at the

same time to draw conclusions on the basis of this information and formulate further

search goals. Intelligent agent – as a concluding system – influenced by external

impulses makes logical, precisely determined decisions. A group of cooperating soft-

ware agents makes up a multiagent system. In accordance with the MAS systems

Kiełtyka L. and Niedbał R. (2005).

A Conception Of Multiagent Management System Of Dispersed Market Information – E-negotiations Area.

In Proceedings of the 2nd International Workshop on Computer Supported Activity Coordination, pages 48-55

DOI: 10.5220/0002559000480055

 SciTePress

pragmatism even already existing information-decision making system can be treated

as environment in which agents carrying out particular tasks are inserted [4].

The Internet has become communication medium and has created possibility of

electronic business management. Its development lets us assume that greater and

greater part of economic life will take place by means of the internet technology.

Economic globalization forces constant communication of market participants and

thus frequent negotiations.

Results of the survey conducted in Polish enterprises were presented in chapter 2.

The aim of the survey was to show whether business negotiations are treated as

a source of market information, and if they are supported by electronic media and

what tasks would Polish managers delegate software agents to perform.

The conception of multiagent system, which is to show possibility of market in-

formation management in an enterprise was proposed in chapter 3.

2 Market Information And Business Negotiations Aided By The

Internet

Business negotiations are the important source of market information acquisition.

Being treated as the communication and information exchange processes, they give

the enterprise opportunity to complete knowledge of the other, negotiating, side [9].

Connection of the contemporary enterprises by means of informatic networks, but

first of all of production and cooperative ties causes that aiding negotiation processes

- perceived as information-decision making processes - becomes an important task of

the enterprise’s informatic system. Business negotiations may take place on the Inter-

net on the virtual B2B (Business to Business) markets, where auctioning mechanism

is not enough. Sides have to use more advanced e-negotiation technologies in order to

fix not only prices but also delivery conditions, technical details, guarantee condi-

tions, payment terms and many more aspects of the agreement. Acquisition of reliable

information is the basis to create loyalty, which in turn means, that there is a possibil-

ity of negotiations in new areas – among other things in the company-client relation-

ships.

Within the confines of the research conducted at Częstochowa University of Tech-

nology, Management Faculty, in the Department of Management Information Sys-

tems, connected with the application of multiagent system and integrators of informa-

tion flow in market information management in the enterprise, there was the survey

“E-negotiations in Polish Enterprises” conducted. The subject of the study were the

enterprises applying in their operations internet business model and mixed business

model – traditional-internet one. The survey was conducted among four groups of

respondents – presidents, members of the board, middle-level managers and employ-

ees. The total number of respondents constituted 172 people. The surveyed enter-

prises included:

− 31 employing more than 500 people,

− 14 employing 251-500 people,

− 44 employing 51-250 people,

− 49 employing 11-50 people,

− 34 employing fewer than 10 people.

The majority of 117 constituted the enterprises with 100% of Polish capital. The

remaining enterprises were:

− 27 with the majority of Polish capital,

Fig. 1. Perception of tasks in Polish companies most willingly assigned for an intelligent soft-

ware agent to perform. Source: own analysis.

− 24 with the majority of foreign capital,

− 4 with 100% of foreign capital.

With respect to the type of operation the biggest group of all the enterprises consti-

tuted service companies – 61, then production companies – 42, trade companies – 31,

production-trade-service companies – 15, trade-service companies – 14, production-

trade companies – 9.

The survey consisted of two parts and included 23 questions altogether that con-

cerned business negotiation aiding and market information acquisition by electronic

media, the Internet in particular. The results of the two questions directed to respon-

dents are presented below. The questions were connected with the perception of the

software agents’ role and business negotiations, as the sources of market information

(Fig. 1, Fig. 2).

It follows from the chart presented in Figure 1 that one of the most burdensome

and important at the same time activities for managers responsible for business nego-

tiations in Polish companies is – according to the respondents – collecting informa-

tion about prices of competitive products (7,6 points in 10-point scale). It is acquisi-

tion of information on prices of competitive products that they would assign an intel-

ligent software agent to perform these duties. Thus, there exists a real need to auto-

mate this activity and if at all possible quick implementation of such a solution. How-

ever, the respondents would be less willing to allow the software agent to select their

negotiating partner (5,1 points).

Fig. 2. Perception of business negotiations functions in Polish companies. Source: own analy-

sis.

The result of the second answer is connected among other things with classical

methods of the other side’s credibility verification. Nevertheless, it should be stressed

here that the respondents assigned substantial significance to every activity repre-

sented in Figure 1. It justifies the necessity of the intense research in the area of soft-

ware agents application in social organizations, especially with regard to market in-

formation acquisition.

The second of the presented in this paper questions concerned perception of the

role of business negotiations in Polish companies. The results received presented in

Figure 2 indicate that the most important role of negotiations is maintaining or win-

ning the competitive advantage (6,6 points). However, the two following positions

are connected with the notion of information. Negotiations are as follows: the tools of

organizing information exchange with the environment and the source of information

of the environment (comparable significances – 5,9).

The most important role in enterprises is still played by traditional “face-to-face”

negotiations where negotiators meet directly. This concerns particularly negotiations

of very important or strategic for enterprises contracts. Undoubtedly, it is still the

most interactive way of communication between the negotiating sides.

However, the analysis of the conducted survey and interviews justify necessity of

work on application aiding managers in running e-negotiations, particularly in rela-

tions on electronic, B2B markets.

3 Proposed Model Of The System

One of the conceptions of a system aiding market information management in the

e-negotiations area is its realization in the form of multiagent system. The proposed in

the work model will comprise of two mutually cooperating groups of software agents

(so called agencies):

− First group – its task is to acquire, analyze and make market information available

to manager, aids manager in preparation to negotiations phase,

− Second group – its task is to aid manager in the terms of transaction negotiations; it

takes over the task of offers evaluation and working out initial compromise.

The above mentioned groups include three basic kinds of agents:

− Interface agents – their task is to facilitate man interaction with the system and

presentation of the work results in accessible form.

− Coordination agents – they analyze introduced by man tasks and divide them into

elementary tasks.

− Information agents – they realize elementary tasks, delegated to them to carry out

by task agents (they take and store data from the available information resources,

collect information acquired in the process of negotiation, take over tasks of offer

exchange and working out initial compromise); they return results of their opera-

tions to Interface Agents.

General architecture of the first group of agents presents Figure 3.

Information agents take over tasks of collecting, analyzing and making market in-

formation helpful in preparation to negotiations with clients and suppliers available to

manager. They also take over tasks of offers exchange – so called packages. Offers,

that is lists of items to negotiate and their weights are sent as KQML (Knowledge

Query and Manipulation Language) language commands. Items to negotiate are

among others: price, quantity, quality, delivery terms and conditions, payment condi-

tions, guarantee, discounts. Packages are subject to evaluation, and worked out by

Information Agents possible compromises are presented to managers to accept or

modify with respect to the weight of particular negotiated items. If the agreement is

accepted, Information Agents cease their operation and the final part of the negotia-

tions concerned with signing contract and accounting is taken over by man. In the

case of lack of acceptation however, Information Agents work out compromise again

or their operating is interrupted. Thanks to application of Information Agents and the

possibility of holistic packages comparison by managers there exists a chance of

Pareto-optimum compromises achievement [9].

The proposed in the work model was elaborated in the AgentBuilder environment

of Acronymics Inc. The research of the simulating model was divided into three

stages:

− formulating, describing and assigning tasks to software agents,

− multiagent system implementation – elaborating system’s ontology, establishing

concluding rules of particular agents, negotiating protocol and communication

with database,

− elaboration of coordination method among agents,

− simulation and verification of the proposed model.

Where:

JDBC Agent - software agent for database service.

HTTP Agent – software agent carrying out function of information acquisition from websites.

E-Mail Agent – software agent for electronic mail service.

FTP Agent – software agent for file transfer service.

Auxiliary Agent – for communication with other software agents in KQML language.

Fig. 3. General architecture of the first group of agents. Source: own analysis.

At present the research focuses on the second stage – MAS system implementa-

tion. Initial simulations concerning contract negotiations among software agents rep-

resenting two sides of negotiations are being carried out – Figure 4.

The criteria for evaluation of negotiated packages is value of their usefulness func-

tion described by the formula (1):

∑∑

ijkij

pUxu

)( .

(1)

Where:

U(p

) – p

package usefulness,

k – number of packages,

i – i item to negotiate,

j – j option for i item to negotiate,

– partial usefulness of j option for i point to settle,

ijk

– binary variable indicating if j option of i item is present in p

package,

m – quantity of items to negotiate,

(i = 1……m) – quantity of vital items to negotiate.

Fig. 4. Transaction terms negotiations – offers evaluation. Source: own analysis.

4 Summary

Acquisition of up-to-date market information is one of the crucial factors influencing

management efficiency of contemporary enterprises

. Enterprises possessing proper,

efficient information will not perceive external environment as a barrier in their de-

velopment, but rather as a stimulus to innovative operations

. Due to this fact a higher

efficiency level of market information management should become a goal to pursue

The fulfillment of this postulate requires a new model of information system to be

designed, characterized by elasticity, autonomy and reactivity to changes occurring in

the enterprise’s environment

. Its monitoring is a process of constant following and

search for market changes, sending offer inquires, etc

. Changes happening in the

enterprise’s environment often generate surplus of information [4]

. Not all the enter-

prises are able to make the proper use of this fact

. Efficient system of market informa-

tion largely conditions creation of the market offer by the enterprise, and also influ-

ences maintenance or gaining of competitive advantage on the market

. Moreover,

enterprise management often requires real-time reaction, which means receiving in-

formation about an economic event the moment it happens and immediate dispatch of

feedback information.

Business negotiations are an important source of market information – which is

clearly visible when analyzing negotiating processes from the perspective of their full

“life cycle”

. Information is generated at various stages of the negotiating process.

The proposed in the paper conception of multiagent system aims to show possibil-

ity of market information management through improvement of enterprise’s business

negotiations with other, present in its environment subjects. This is particularly im-

portant because there are often multisided negotiations run in enterprises.

„The research financed from the funds of The State Committee for Scientific Re-

search in the years 2004-2006 as the research – 1H02D0727”

References

1. AgentBuilder. An Integrated Toolkit for Constructing Intelligent Software Agents. User’s

Guide. Version 1.4 Rev.0. Acronymics, Inc. Arizona, Mesa (2004)

2. Chaves, I., Monteiro, E.: MATE – Mobile Agent Technology for Ecommerce. Proceedings

of the International Association for Development of the Information Society International

Conference e-Society 2003. Volume II. Lisbon, Portugal (2003) 730-734

3. Jennings, N. R., Wooldridge, M. J.: Applications of Intelligent Agents. Agent Technolo-

gies: Foundations, Applications, and Markets. Springer-Verlag: Heidelberg, Germany

(1998)

4. Kiełtyka, L., Kucęba, R., Niedbał, R.: Coordination of information processes of the con-

temporary enterprise. Economic informatics. Scientific-Didactic Review No 1027.

Scientific editors : Goliński J., Jelonek D., Nowicki A. Published by Wydawnictwo

Akademii Ekonomicznej im. Oskara Langego we Wrocławiu, Wrocław (2004) 20-26

5. Kiełtyka, L.: Communication in management. Technology, tools and forms of information

transfer. Agencja Wydawnicza PLACET, Warszawa (2002)

6. Kersten G., Kersten M., Noronha S.: Computer aiding and research of negotiating proc-

esses. Zagadnienia Naukoznawstwa, Vol. 136, No. 2 (1998) 202-227

7. Kiełtyka, L.: Multimedia in business. E-commerce, multimedia information technology,

information management, multimedia in teaching. Scientific editor L. Kiełtyka. Kantor

wydawniczy ZAKAMYCZE, Kraków (2003)

8. Kisielnicki, J., Sroka, H.: Business information systems. Informatics in management.

Methods of system designing and systems implementation. Agencja wydawnicza PLACET,

Warszawa (1999)

9. Rządca, R.: Negotiations in business. The way the organizations negotiate. PWE,

Warszawa (2003)

10. Sycara, K.: Multiagent Systems. http://www.aaai.org/Resources/Papers/AIMag19-02-

007.pdf (1998)

11. Várkoly, L.: Possibilities of PC and computer networks application in the selected eco-

nomic departments. Trencianska univerzita v Trencine, Trencin (2001) (in Slovak)