Akademik Veri Yönetim Sistemi
37th International Conference on Computers and Industrial Engineering 2007, Alexandria, Mısır, 20 - 23 Ekim 2007, cilt.1, ss.405-410
The paper is based on developing two data mining algorithms and using them together in order to get smart decisions from large databases. This subject will be exemplified by a system which finds effectively particular products that are mostly sold together regarding customer profiles. Customer profiles become crucial since it makes possible finding customers who are most likely to purchase the product. The algorithms are Clustering Algorithm based on customer profiling and Association Algorithm based on market basket analysis. Our algorithms were constructed on customer transaction history data. The process of mining large databases purposes finding new knowledge, hidden predictive information and presenting them in understandable manner. The efficiency has been increased by using these two algorithms together. We believe that our system constitutes a test bed for the analysis and development of data mining techniques for active customer profiling to make inferences for marketing campaigns. At the end, the results have been accessed on the web. The application, which based with Web technologies, is very useful for getting information from different locations and systems. Web Technologies can share all locations for the people, who have web browser in their own PC. So, a manager can get smart decision on web wherever he is. The application has been programmed by using ASP.NET with C#. programming language based on SQL 2000 database engine.