欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
本系统基于模糊联想记忆神经网络，建立偏好评价模型，根据用户偏好对搜索引擎搜索到的候选文献进行评级，为用户推荐偏好值高的文献。本系统的学习模块采用PCA—CG算法和误差反向传播算法，以用户阅读过的基准文献和其对应评级作为训练样本，对用户偏好进行学习；推理模块根据学习到的模糊规则和隶属函数来计算候选文献的偏好值，并以偏好值排序，把偏好值高的文献推荐给用户。把该模型应用于信息技术类文献的检索，实验表明系统提供的推荐文献具有较高可信度。
英文摘要：
A preference evaluation model is constructed based on Fuzzy Associative Memory Neural Network. It＇ s proposed to recommend literatures with high preference values to users among literatures retrieved by search engines and ranked according to user＇ s preference. The system＇ s learning module uses PCA - CG algorithm and error hack propagation algorithm to learn user＇ s preference with the reference literatures read by users and their corresponding appraisal as training samples. The reasoning module computes the preference values according to fuzzy rules and membership functions, ranks the result from high preference to low preference and recommends the literatures with high preference to users. The model is applied in the retrieval of literatures of information technology category, and the experimental result proves that the literatures recommended by the system is reliable.
同期刊论文项目
同项目期刊论文
期刊信息
位置：