共查询到19条相似文献,搜索用时 187 毫秒
1.
2.
3.
4.
项武 《微电子学与计算机》2010,27(7)
研究协同过滤推荐系统处理大规模稀疏评价数据的精度问题,针对目前余弦相似性处理极稀疏的用户评价矩阵不能获取满意推荐质量,且处理大规模推荐数据时存在性能缺陷的情况,提出基于评分区间相似性的协同过滤推荐算法.将用户评价的数值范围进行定性划分,以项目在各评价区间上的得分建模项目质量,以用户在各评价区间内的评分建模用户评分取向.使用余弦相似性度量方法衡量用户和项目间的相似程度,进而确定最近邻居.此方法的实现过程降低了稀疏性对最近邻发现的影响,且算法具有快速处理大规模评价数据的能力.使用标准的MovieLens数据集,测试提出的推荐算法,实验结果表明,本算法相比传统的最近邻推荐算法具有更好的推荐精度. 相似文献
5.
6.
由于新用户和新项目的不断加入,传统的协同过滤推荐算法存在冷启动问题。针对该问题,提出了一种改进相似度计算方法的协同过滤推荐算法。首先根据项目的属性特征,计算项目的属性相似性,然后根据项目的用户评分,计算项目的得分相似性,按一定的权重比例将两种相似性组合起来作为最终的项目相似性。最后,根据项目相似性计算目标项目的邻居项目集,根据邻居项目集预测目标项目的用户评分。实验结果表明,新算法能提高推荐精度,并能在一定程度上解决冷启动问题。 相似文献
7.
8.
为了克服协同推荐系统中的用户评分数据稀疏性和推荐实时性差的问题,提出了一种高效的基于粗集的个性化推荐算法.该算法首先利用维数简化技术对评分矩阵进行优化,然后采用分类近似质量计算用户间的相似性形成最近邻居,从而降低数据稀疏性和提高最近邻寻找准确性.实验结果表明,该算法有效地解决用户评分数据极端稀疏情况下传统相似性度量方法存在的问题,显著地提高推荐系统的推荐质量. 相似文献
9.
针对协同过滤推荐中存在的数据稀疏性问题,文中提出了一种基于项目特征属性和BP神经网络相结合的协同过滤推荐算法,并通过Movielens数据集验证了该模型的有效性。此方法首先将用户评分数据映射为用户对项目特征属性的偏好,然后使用BP神经网络训练得到目标用户的特征属性偏好模型并对新项目的评分进行预测,从而降低用户项目评分矩阵的数据稀疏性,最后使用协同过滤推荐算法,形成最近邻并生成推荐建议。 相似文献
10.
针对数据稀疏性问题,提出基于蚁群聚类的项目评分预测方法.在对Web日志分析基础上将用户聚类,针对目标用户的未评分项目,找到目标用户的若干最近邻类簇,利用类簇内其他用户对目标项目的评分预测未评分项目的评分,从而达到降低数据稀疏性目的.最后,结合协同过滤思想设计了相应的推荐算法,并用从自主开发的旅游电子商务网站上收集的数据进行试验仿真.实验结果表明,与其它缓解数据稀疏性的方法相比,文中的方法显著提高了推荐精度. 相似文献
11.
12.
13.
基于k-近邻的协同过滤推荐算法对于邻居数量k的确定过于主观,并且推荐时以k-近邻均值加权推荐不够准确.针对这两个问题,本文首先引入并改进最大最小距离聚类算法,进而设计启发式聚类模型将用户进行不规定类别数的自由聚类划分,目标用户所在类的用户为邻居用户,客观确定邻居数量;然后在推荐时定义类别相似度,针对性地建立目标用户未评分和评分项目的潜在类别关系,改进k-近邻均值加权算法.实验结果表明,该算法提高了推荐准确度(约0.035MAE). 相似文献
14.
To address the problem that in a practical recommendation system (RS),because of the datasets are often very sparse,the traditional collaborative filtering (CF) approach cannot provide recommendations with higher quality,a novel CF based on rough set rule extraction was proposed.Firstly,the attributes of user/item and the user-item rating matrix were used to construct a decision table.Then,the core value of each rule in the table was extracted through using the decision table reduction algorithm.Finally,according to the nuclear value decision rule of the core value table,the reductions of all decision rules were utilized to predict the rating scores of un-rated items.Experimental results suggest that the proposed approach can alleviate the data sparsity problem of CF,and provide recommendations with higher accuracy. 相似文献
15.
《中国邮电高校学报(英文版)》2014
Collaborative filtering algorithms have become one of the most used approaches to provide personalized services for users to deal with abundance of information. The traditional algorithms just use the explicit user-item rating matrix to find similar users or items. To improve the accuracy of the ratings predicted by the collaborative filtering algorithms, social information is widely incorporated into the traditional ones. Different with the existed works focus on directly connected neighbors, we consider the community between the users. We design the algorithms in two aspects: one is that the members in the same community have similar tastes and preferences, the other is that the member's taste is affected by the other members. We simplify these two factors as community similarity and community affection. Community similarity is incorporated into modifying the model-based collaborative filtering algorithm as the social community-based regularization (SCR), which improves 6.2% in mean absolute error (MAE) and 6.1% in root mean square error (RMSE) compared to the existed social recommendation algorithm. Community affection is incorporated into modifying the neighborhood-based collaborative filtering algorithm as the neighbor-based collaborative filtering based on community detection (NCFC) which improve 14.8% in MAE and 8.1% in RMSE compared to user-based collaborative filtering (UCF). 相似文献
16.
计算相似度时,协同过滤算法会赋予所有用户或物品一致的相似度权重,进而导致相似度计算出现偏差。针对这一问题,文中提出一种改进相似度的协同过滤算法。该算法首先在计算用户间相似度时根据用户活跃量增加活跃用户惩罚因子,然后在计算物品间相似度时根据物品流行度增加热门物品惩罚因子,再对相似度做最大值归一化,最后根据相似度矩阵进行电影评分预测。实验结果表明,改进的相似度算法在评分预测时更加准确,平均绝对误差稳定在0.72左右。 相似文献
17.
旅游业和网络时代高速发展,导致旅游信息过载问题日益严重,旅游推荐方法对解决信息过载问题十分重要。传统推荐算法只针对用户和项目之间的评分和基本属性计算相似度进行推荐,但行为需求及具有游客情感因素的评论却被忽视。本文利用卷积神经网络(CNN)对文本评论特征提取进行情感分类,用皮尔逊相似度公式计算相似的用户群体,用平均绝对误差(MAE)对结果误差进行评价。与传统的协同过滤方法进行对比,本文提出的模型能有效降低预测误差。 相似文献
18.
POI (point of interest) recommendation is an important personalized service in the LBSN (location-based social network) which has wide applications such as popular sights recommendation and travel routes planning.Most existing collaborative filter algorithms make recommendation according to user similarity and location similarity,they don’t consider the trust relationship between users.And trust relationship is helpful to improve recommendation accuracy,robustness and user satisfaction.Firstly,the propagation property of trust and distrust relationship was analyzed.Then,the measurement and computation method of trust were given.Finally,a hybrid recommendation system which combined user similarity,geographical location similarity and trust relationship was proposed.The experiments results show that the hybrid recommendation is obviously superior to the traditional collaborative filtering in terms of results accuracy and user satisfaction. 相似文献
19.
针对数字图书馆图书借阅推荐系统中通常采用的协同过滤算法中存在的不足,分析了该算法存在在用户多兴趣情况下推荐时难免会产生偏差,从而降低预测推荐准确率问题,进而提出了一种基于用户多兴趣度的图书借阅推荐系统推荐算法的设计思路。该算法首先对待推荐图书进行分类,然后根据用户的借阅情况计算其对每类图书的兴趣度,将对每类图书兴趣度相似的用户聚为一类,在小范围内寻找最近邻居,最后根据用户对不同类别书目的兴趣度作为权重值来产生对该用户的每类图书的推荐数目。该算法不仅能够解决用户多兴趣度问题,同时也可减少计算工作量,提高推荐的准确率。 相似文献