Research on e-shopper purchase behavior based on linguistic values of attributes of database

It is very necessary for e-marketers to understand e-shoppers' needs for making correct marketing strategies. Knowledge of e-shoppers purchase behaviors can be extracted from transaction databases. In this paper, transaction database is transformed into the database of linguistic values, and research on e-shopper purchase behavior is made based on linguistic values of product attributes. And an algorithm that contains the first support, the second support and rel-confidence is proposed to extract rules from the database of linguistic values. The algorithm is applicable for mining both frequent linguistic items and significant rare items. Consequently, e-shoppers purchase behavior can be revealed based on the rules. The result indicates the proposed method is very efficient in discovering e-shoppers purchase behavior, and is useful for e-marketers to make marketing decisions.