For the Shipbuilding Enterprise Model of Intelligent Recommendation Services

Based on ships transaction in useful information personalization gain question, proposed that one kind the intelligent recommendation service model which unifies the ASP (Application Service Providers) pattern and the recommendation technology. In this model, the service uses five structures, the intelligence management transfer application service level and the resources component level, cooperates to realize the intelligence mutually. In view of the ships profession electronic commerce pattern's characteristic, the depth limited auto-adapted k close neighbor searching algorithm which uses transforms the user grading non-isotropic space as the isotropic space, obtains the isotropic grading matrix, thus searches the current user k recent neighbor, has current user's forecast grading, and has the recommendation. Has provided the simple direct-viewing practical commercial service for the ships manufacturing firm.