本体技术在个性化推荐系统中的应用研究

Application of ontology technology in personalized recommendation system

针对个性化推荐系统中传统的用户建模技术的缺陷,本文将领域本体的理论与方法应用到电子商务个性化推荐中,设计了一种本体和服装信息相结合的个性化推荐模型。利用Protégé实现了服装知识本体的构建过程,在真实数据集上的实验结果验证了算法的可行性和有效性。该设计与已有的推荐算法相比,具有更高的预测准确度。

In view of the negatives of conventional user modeling technology in personalized recommendation system, we apply the theory and method of domain ontology to personalized recommendation of e-commerce, and design an ontology technology and clothing information combined recommendation model. We implement the construction procedure of clothing ontology with ontology editor tool Protege. Feasibility and effectiveness of the model are verified on practical datasets. It has higher prediction accuracy, as compared with existing recommendation models.