基于改进蚁群算法的多媒体图像个性化推荐模型  

Personalized recommendation model of multimedia image based on improved ant colony algorithm

作　　者：陈钊 沈舒海 CHEN Zhao;SHEN Shu-hai(Anhui Vocational and Technical College,Hefei 230011,China)

机构地区：[1]安徽职业技术学院,安徽合肥230011

出　　处：《齐齐哈尔大学学报（自然科学版）》2021年第4期44-48,共5页Journal of Qiqihar University(Natural Science Edition)

摘　　要：为了提高多媒体图像个性化推荐能力,提出基于改进蚁群算法的多媒体图像个性化推荐模型。构建多媒体图像的边缘像素特征解析模型,采用特征空间重组实现对多媒体图像的像素结构重组,采用多级语义特征分析方法进行多媒体图像个性化信息融合和特征图融合处理,采用浅层特征图组合控制方法实现对多媒体图像个性化特征提取,对提取的多媒体图像个性化特征采用蚁群算法进行寻优参数识别,构建多媒体图像个性化推荐的多层次特征解析模型,结合多媒体图像的深层特征信息分布进行个性化推荐。仿真结果表明,采用该方法进行多媒体图像个性化推荐的使用度水平较高,推荐精度较好,提高了多媒体图像个性化融合和特征辨识能力。In order to improve the personalized recommendation ability of multimedia images,a personalized recommendation model of multimedia images based on improved ant colony algorithm is proposed.The edge pixel feature analysis model of multimedia image is constructed,the pixel structure of multimedia image is reorganized by feature space reorganization,the personalized information fusion and feature map fusion of multimedia image are processed by multilevel semantic feature analysis method,the personalized feature extraction of multimedia image is realized by shallow feature map combination control method,the extracted personalized features of multimedia image are identified by ant colony algorithm,and the multi-level feature analysis model of personalized image recommendation of multimedia image is constructed.Simulation results show that this method has higher usage level and better recommendation accuracy,and improves the ability of personalized fusion and feature recognition of multimedia images.

关 键 词：改进蚁群算法 多媒体图像 个性化 推荐 特征融合 

分 类 号：TP391.3[自动化与计算机技术—计算机应用技术]