With the rapid growth of internet and multimedia technology, people have been recently interested in Web-based electronic commerce, and buyers tend to request the special personalized services. However today most the conventional systems for electronic commerce only provide the catalog of goods which contains the shapes and the standards of goods.
Required is a sale agent which can offer the information about goods more quickly and conveniently by using sale clerks' knowledge beyond the level of service only offering the information about goods to satisfy the demand of buyers. The use of this kind of sale clerk agent makes buyers save the time for searching for goods and do shopping that is appropriate for buyers' taste.
In this thesis, we design and implement an ISA(Intelligent Sale Agent) for cyber shopping mall, in which buyers can buy the goods for their taste, in stead of sale clerks in the actual shops. Threshold, one for fetching goods which buyers randomly choose according to buyers' taste, and fuzzy theory for considering effectiveness and efficiency of goods retrieved, can increase the correctness of the list of goods. After buyers buy goods, the preference level of buyers is gotten and stored in buying pattern database.
With multiple regression considering the stored buying patterns and the profile of customers, only a variable affecting the propensity to buy goods is withdrawn and the buying patterns are learned using the neural network. After learning, new buyers are guided with the learned pattern information using the neural network.
In case of the existing ones, the old buying pattern material is used for their buying of the same goods as they bought and they are guided to the goods using the learned materials after buying the new goods. we have estimated the buying pattern of buyers after making the old buying patterns learned using neural network and made the ambiguous buying attributes of buyers expressed using fuzzy theory. Therefore the shopping mall can satisfy the buyers' desire and buyers can buy the goods suiting their taste quickly because of the structure of shopping mall using ISA.