Item folksonomy or tag information is a kind ofudtypical and prevalent web 2.0 information. Item folksonmyudcontains rich opinion information of users on itemudclassifications and descriptions. It can be used as anotherudimportant information source to conduct opinion mining. Onudthe other hand, each item is associated with taxonomyudinformation that reflects the viewpoints of experts. In thisudpaper, we propose to mine for users’ opinions on items basedudon item taxonomy developed by experts and folksonomyudcontributed by users. In addition, we explore how to makeudpersonalized item recommendations based on users’ opinions.udThe experiments conducted on real word datasets collectedudfrom Amazon.com and CiteULike demonstrated theudeffectiveness of the proposed approaches.
展开▼