数字图书馆个性化服务用户模型研究

提出了一种数字图书馆个性化服务用户模型构架,并对实现过程中的几个关键问题,包括用户模型表示方法,用户模型的建立以及更新算法进行了详细论述.提出了基于本体论的空间向量用户模型表示方法,建立了简单的数字图书馆领域本体,以空间向量表示用户模型,以本体概念作为向量的特征项;采用支持向量机分类算法和无监督聚类算法相结合提取用户兴趣;在用户模型更新方法上,采用渐进遗忘和滑动窗口相结合的方法实现用户兴趣概念的漂移.

Study on the User Profile of Personalized Service in Digital Library

Structure and mechanism of user profile in digital library personalized service is presented.Key problems about the construction of user profile are introduced in details. These problems include the representation, creation and updating of user profile. The user profile is represented by a vector space model (VSM) based ontology. Simple domain ontology of digital library is built up first, and then the user's interests represented by VSM, in which the item of vector is ontology or concept about user's interests. User's interests are extracted by combining classification and clustering algorithms, taking the support vector machine (SVM) and unsupervised clustering (UC) as examples to show their implementation. Updating of user profile is implemented by using hybrid algorithm, containing time window and gradual forgetting.