This paper focus on analysis and implementation of a hash table based data mining engine used to discover knowledge from transaction database for electronic commerce industry. Apriori algorithm is analyzed and implemented to discover association relationship among amounts of business transaction records. Such relationships can be derived based on individual product purchased in all transactions. They also can be applied to single category level that will provide more granular analysis. This engine can derive both single and multidimensional association rules. The resulted association relationships help in much business decision-making processes, such as catalog design, shelf layout, cross-marketing, bundled advertisement, etc.