摘要
针对目标用户所感兴趣的图书具体特征,开发了一种基于意见挖掘技术结合图分类器的图书推荐系统.首先,基于图的分类器技术对图书条目进行分类;然后,挖掘出图书条目的评价信息,并根据评价内容判定评价的正反面;最后,结合两组技术实现图书的个性化推荐.通过实际案例的数据进行计算,以精确率和召回率评价依据进行综合评价,结果证明了提出的推荐系统的可行性和实用性.
In view of the specif ic features of target users interes ted in books, a book recommendation systern is developed based on opinion mining technology. Fi rst of al l , based on the f igure of the classif ier technology for classif icat ion of book ent ries; Then, excavated book entries of assessment information, and according to the posit ive and negative evaluat ion content determine evaluat ion; Final ly, combining wi th two sets of techniques to realize personal ized books recommend. Through the data of actual case is calculated to precision rate and recal l rate evaluation on the basis of comprehensive evaluat ion, the results prove the feasibility and practicabi lity of the recommendation system is put forward.
出处
《湘潭大学自然科学学报》
北大核心
2017年第3期107-110,共4页
Natural Science Journal of Xiangtan University
基金
海外及港澳学者合作研究基金项目(61028003)
关键词
意见挖掘
图分类器
图书推荐系统
贝叶斯
opinion mining
figure classi fier
book recommendation s ystem
Bayes classifier
