In order to provide personalized suggestions for consumers in a more accurate way, Intelligent Recommendation System (IRS) has emerged. This study made a meaningful exploration on its structure and its implementation. By combining the fuzzy clustering, fuzzy association rules mining and fuzzy reasoning, this study designed and implemented an original IRS. It could give more matching suggestions based on the authentic data and the operation of fuzzy clustering, fuzzy associations rules mining, rules filtering, and fuzzy reasoning.