论文部分内容阅读
文章提出了一种基于本体的用户建模方法。首先,基于用户兴趣概念集合,分别从同主题源本体的RDFS闭包中抽取子本体,然后采用多策略映射方法将其合并形成初始用户本体;其次,基于各类用户Web使用数据,挖掘高频关键词,追踪用户兴趣变化,构建增量本体;然后,利用衰减函数和相关阈值,将增量本体与初始用户本体合并,在增加新概念及其关系的同时,剔除不再感兴趣的概念,以实现用户本体的自动优化。
The article proposes a method of ontology-based user modeling. Firstly, based on the set of user interest concepts, the sub-ontology is extracted from the RDFS closures of the subject source ontology respectively and then combined by the multi-strategy mapping method to form the initial ontology. Secondly, based on the data of various users’ Web usage, Key words to track changes in user interest and build incremental ontology; then, using the decay function and the correlation threshold, the incremental ontology is merged with the initial user ontology, while adding new concepts and their relations, the concept of no longer interested is removed, In order to realize the automatic optimization of the user’s body.