Optimal Tasks Assignments in a Large Accountancy Firm Using Dialogues With Fuzzy Expert Systems

Optimal Tasks Assignments in a Large Accountancy Firm Using Dialogues With Fuzzy Expert Systems

conference paper

Czech

Abstract. A large accountancy firm employs several tens of accountants and gets hundreds of accountancy tasks of different sizes and nature. Each task has to be assigned to an accountant. In theory this problem can be solved by well known methods of operation research. However, some very important characteristics of different accountants are very vague and therefore difficult to quantify e.g.: Creativity, Loyalty, Ability to gain confidence of clients. The mentioned characteristics make applications of traditional e.g. scheduling algorithms prohibitively difficult. Moreover some accountancy tasks are unique and their assignments can be done just by an experienced manager and not by an algorithm. Fuzzy descriptions of tasks and accountants are studied in this paper. A two level assignment process is studied. The preliminary screening decides if the task under study is a routine task or not. The routine task is such task which can be assigned by the fuzzy algorithm. A non-routine task assignment is made by a dialogue: manager ? fuzzy expert system. The case study presents a fuzzy knowledge base with 30 dimensional fuzzy descriptions of 40 accountants. The knowledge base is used to make a routine assignment and a short dialog represented by a sequence of queries and answers is described in detail.

Abstract. A large accountancy firm employs several tens of accountants and gets hundreds of accountancy tasks of different sizes and nature. Each task has to be assigned to an accountant. In theory this problem can be solved by well known methods of operation research. However, some very important characteristics of different accountants are very vague and therefore difficult to quantify e.g.: Creativity, Loyalty, Ability to gain confidence of clients. The mentioned characteristics make applications of traditional e.g. scheduling algorithms prohibitively difficult. Moreover some accountancy tasks are unique and their assignments can be done just by an experienced manager and not by an algorithm. Fuzzy descriptions of tasks and accountants are studied in this paper. A two level assignment process is studied. The preliminary screening decides if the task under study is a routine task or not. The routine task is such task which can be assigned by the fuzzy algorithm. A non-routine task assignment is made by a dialogue: manager ? fuzzy expert system. The case study presents a fuzzy knowledge base with 30 dimensional fuzzy descriptions of 40 accountants. The knowledge base is used to make a routine assignment and a short dialog represented by a sequence of queries and answers is described in detail.

Keywords: fuzzy, fuzzy expert system, multi-objective decision-making, accountancy firm, artificial intelligence

Keywords: fuzzy, fuzzy expert system, multi-objective decision-making, accountancy firm, artificial intelligence

KÁBA, D.; KORÁB, V.

2009

29. 10. 2009

University of Seville

Seville, Spain

978-84-692-6077-7

Innovation & Change in European Environment

1

1

48

57

120