Informing Science InSITE- “Where Parallels Intersect ” Jun

Credits are the one of the most important functions in bank management, because, from one side, thanks to a good credit policy a bank can earn money, but from the other side, in the case of weak or wrong credit policy the bank can make substantial losses.

Informing Science InSITE - “Where Parallels Intersect” June 2002

TAL Rules versus ECA Rules: an Attempt for Comparison in

the Credit Management Context

Maria Mach

University of Economics, Wroclaw, Poland

Abstract

Credits are the one of the most important functions in bank management, because, from one side, thanks

to a good credit policy a bank can earn money, but from the other side, in the case of weak or wrong

credit policy the bank can make substantial losses. Therefore in the field of credit policy management,

intelligent information systems can be very helpful, as it is a complex and heterogeneous field, needing

complex management and decision-making procedures.

There exist many technical solutions aimed at helping the decision-makers in this field, from “traditional”

ones, as databases, to more sophisticated tools, as for example expert systems, the main aim of which is to

perform the analysis of applications for a credit, thus helping to make proper credit decisions.

Credit management is closely related to time, in other words, the temporal aspect of credit management

can be very clearly seen. Therefore while building intelligent systems in this area, it would be recom-

mended to take this temporal aspect into account.

The article concentrates on the question of searching and choosing an intelligent computer tool which

would fulfil the above mentioned requirements, the toll which would help to make necessary credit analy-

ses, to make proper credit decisions, taking into account the temporal aspect of credit management. Two

solutions are discussed: TAL language and active databases. Some exemplary credit management rules

are encoded both in the TAL language and in the form of ECA rules. Both kinds of rules are analysed and

discussed, as well as their advantages and disadvantages.

Keywords: Credit management, TAL rules, active databases, ECA rules, intelligent systems

Introduction

Credits constitute one of the most important functions in bank management process, a function which

very often decides about bank’s “to be or not to be”. It is therefore clear that credit policy is such a field,

in which computer tools are needed and useful. The attention must also be paid to the fact, that banking

policy is a very complex and heterogeneous field, needing complex management procedures.

There are many concepts and ideas concerning credit management, arising from both computer systems

sensu stricto as well as from intelligent systems. It

is proposed, for example (see e.g. (Ziemba, 1992)) to use expert systems which perform the analysis of applications for a credit, thus helping to make proper credit decisions. Such systems are mainly aimed at minimising the percentage of so-called bad credits.

Credits are the one of the most important functions in bank management, because, from one side, thanks to a good credit policy a bank can earn money, but from the other side, in the case of weak or wrong credit policy the bank can make substantial losses.

TAL Rules Versus ECA Rules

Credit management is closely related to time, in other words, the temporal aspect of credit management

can be very clearly seen. Therefore while building intelligent systems in this area, it would be recom-

mended to take this temporal aspect into account.

In (Mach, 2000) the Temporal Action Language (TAL) was analysed as a tool for performing effective

credit management. While analysing the rules encoded in this language, one can notice immediately their

similarity to the rules of so-called active databases, namely ECA rules. Therefore it is worth answering a

question, whether searching for new tools (such as the TAL language) makes sense, if there exist well-

known and good solutions (active databases). To answer the above question, the rules encoded in the TAL

language will be analysed along with the same rules encoded in the form of ECA rules. The advantages

and disadvantages of both solutions will also be discussed.

The rest of the paper is organised as follows: in section 2 the temporal aspect of credit management is

shortly presented. Section 3 is devoted to main features and ideas of the TAL language, while the second

part of this section is devoted to a short survey of active databases and ECA rules concepts. Section 4 con-

tains examples of credit rules encoded in the TAL language and as ECA rules. The comparison of both

solutions is done in section 5, along with the discussion concerning advantages and disadvantages of both

types of rules. Section 6 contains conclusions and future research suggestions.

The Temporal Aspect of Credit Management

The temporal aspect of credit management can be seen already in the definition of credit. In (Bogacka-

Kisiel, Karwowski, 1991) for example, the authors give the following definition:

“Credit – an economical relation, in which one party (the lender) transfers a predetermined value in

money or commodity to the other party (the borrower), on condition of returning its equivalence with

eventual remuneration (interest) in a predetermined period”.

Of course the question of a term (or terms) does not have to be restricted only to the aspect of returning

the credit, it can also concern the time-ordering of particular credit tranches (Krzy kiewicz, 1996), (Go-

spodarowicz, Mo aryn, 1998), (Jaworski et al., 1997). The above questions are set up in detail …… 此处隐藏：27833字，全部文档内容请下载后查看。喜欢就下载吧 ……