Martinovska, Cveta and Teohareva Filipova, Biljana (2012) Analyzing customer spending patterns and buying preferences with data mining techniques. In: First International Conference for Business, Economy and Finance -I CBEF 2012 - From liberalization to Globalization, 13-15 Sept 2012, Stip, R.Makedonija.
Full text not available from this repository.Abstract
Apart from using information systems for management of procurement data, sales
and purchase transactions, warehouse operations, finance and human resources,
information technologies are increasingly used for analysis, planning and control of business
processes. In contemporary economy companies need various business intelligence
techniques for analyzing business data, such as sales revenue or production costs.
The term business intelligence (BI) was first introduced by Howard Dresher, Gartner
Group analyst, to denote “concepts and methods to improve business decision making by
using fact-based support systems”. Business intelligence as it is understood today provides
tools for online analytical processing, data mining, process mining, business performance
management and predictive analytics. Numerous data mining methods are used for
marketing, sales and customer support: market basket analysis, clustering, neural networks,
decision trees, genetic algorithms, association rules, statistical methods, etc.
There are a lot of programming tools for data mining present on the market,
produced by leading software companies. For example tools which are part of the statistical
program packages, like Enterprise Miner (SAS) and Clementine (SPSS), specialized tools
for general or business usage, such as Intelligent Miner (IBM) and Data Miner (SAS), OLAP
tools, as Hiperion, Pentaho and IBM Cognos Business Intelligence.
Some DBMS include data mining tools, as for exmple Microsoft SQL Server
Business Intelligence and Oracle Data Mining suit Darwin. Besides the above mentioned
tools, there are many others on the market such as: Advanced Miner, Affinium Model,
DataDetective, DataLab, Kalidara Advisor, XLMiner and open source data mining systems
as WEKA, Orange, Tanagra, Rapid Miner, KEEL, KNIME, MiningMart, MLC ++.
In this paper we are considering the implementation of methodologies for market
segmentation, discovering the profile of typical customers for particular kind of products,
their buying preferences and cross selling motivators. We use several data mining
techniques, such as Bayesian network, decision tree and neural network, to determine the
factors that affect the spending pattern and buying preferences of the customers.
From the obtained results several conclusions can be drawn. Customer’s income
affects the spending pattern compared to other factors, such as age, marital status, and the
number of children. Issuing loyalty cards combined with promotional activities and discounts
for pairs of associated products can increases the profit. Using decision tree and market
basket analysis the spending patterns of the customers with loyalty cards were discovered.
| Item Type: | Conference or Workshop Item (Paper) |
|---|---|
| Subjects: | Natural sciences > Chemical sciences |
| Divisions: | Faculty of Computer Science |
| Depositing User: | Cveta Martinovska Bande |
| Date Deposited: | 03 Sep 2013 14:20 |
| Last Modified: | 03 Sep 2013 14:20 |
| URI: | https://eprints.ugd.edu.mk/id/eprint/3603 |
Actions (login required)
 |
View Item |