A Design Service Recommendation Method for Considering User Preference under Cloud Service Platform
-
摘要: 为提高云服务平台下产品设计服务推荐的质量，增强协同产品设计过程中用户与服务提供方之间的交互，构建了个性化产品设计服务推荐框架，并提出考虑用户需求偏好的设计服务推荐方法。采用模糊C均值聚类算法实现用户初步聚类；利用改进的基于用户的协同过滤方法优化用户相似度函数及评分预测函数，以提高目标用户对设计服务推荐项目预测评分的准确性；通过用户对设计服务推荐结果的交互反馈，完成产品设计服务的更新迭代。以设计云服务平台上用户对壁挂式智能空调设计的需求为例，验证了所提方法的有效性。Abstract: Under the cloud service platform, in order to improve the quality of product design service recommendation and enhance the interaction between users and service providers during the process of product collaborative design, the framework of personalized product design service recommendation was built, and a design service recommendation method for considering user demand preference was developed. Firstly, Fuzzy C-means (FCM) clustering algorithm was adopted to fulfill initial clustering of users. Then, optimizing the function of user similarity and score prediction by improved collaborative filtering method based on user (User-based CF) to increase the accuracy of the target users for predicting scores design service recommendation items. Finally, the update and iteration of product design services could be realized through the interactive feedback of users on design service recommendation results. The design requirement of wall-mounted intelligent air conditioning which submitted by the users in Design Cloud Service Platform was taken as an example to verify the effectiveness of the proposed method.
-
表 1 平台用户分类及描述
编号 用户类型 用户描述 用户特征 1 普通
用户通过云服务平台发布设计需求，可与设计师、制造商等实时在线交流，参与协同产品设计制造全过程，获得完整的设计服务解决方案。 非领域性、非
专业性2 有设计
能力的
用户通过云服务平台获得设计知识、资源，上传初步或完整的创意设计方案，与设计师、制造商在线交流，参与协同产品设计制造全过程。 领域性、
专业性或
非专业性
下载: 导出CSV
表 2 设计云服务平台用户信息
用户信息分类 各信息子项 描述词 基本信息I1 性别I11 男\女 年龄群I12 青年\中年 产品属性
偏好I2功能I21 主导功能、辅助功能、
兼容功能结构I22 几何特征、零件构成、
技术约束材质I23 材料、物理特性、
化学特性色彩I24 主色彩、辅色彩、
色彩风格外观I25 造型、质感、
视觉其他I26 智能化、个性化、
交互体验等浏览信息I3 历史浏览行为I31 有\无 信息搜索I32 搜索关键词
下载: 导出CSV
表 3 FCM聚类算法样本对聚类中心隶属度损失函数模型各参数设置
名称 数值 样本点个数i 100 初始聚类个数j 2 加权指数m 2 迭代次数s 100 迭代停止阈值ε 10−6
下载: 导出CSV
表 4 部分用户聚类结果
用户编号 距离 聚类簇编号 用户编号 距离 聚类簇编号 1 6.00 2 11 26.52 4 2 9.92 3 12 27.56 4 3 3.06 2 13 12.60 3 4 9.92 3 14 3.42 2 5 27.56 4 15 15.60 3 6 0.12 1 16 9.30 3 7 21.62 4 17 2.10 1 8 3.80 2 18 3.42 2 9 15.60 3 19 1.56 1 10 9.92 3 20 9.30 3
下载: 导出CSV
表 5 设计云服务平台用户聚类分析结果
类别
序号个人基本
信息产品偏好主要关键词 浏览
信息1 男性，青年 功能、结构、外观、智能化、
个性化、交互体验有 2 男性，中年 功能、结构、材质、智能化 无 3 女性，青年 功能、色彩、外观、智能化、
个性化、交互体验有 4 女性，中年 功能、材质、色彩、外观、
交互体验无
下载: 导出CSV
表 6 用户相似度部分计算结果
用户 传统方法sim /% 改进方法sim′ / % 用户1，用户2 90.65 91.87 用户3，用户4 88.42 89.85 用户5，用户6 87.03 88.83 用户7，用户8 84.27 85.62 用户9，用户10 91.75 92.61 用户11，用户12 90.24 91.70 用户13，用户14 85.56 84.58 用户15，用户16 95.42 95.55 用户17，用户18 90.30 91.05 用户19，用户20 91.11 91.74
下载: 导出CSV
表 7 不同用户评分预测函数得到的RMSE值对比
目标用户
编号RMSE值 改进方法用户评分预测
准确性提高比例 / %传统方法 改进方法 1 0.6463 0.6175 4.46 2 0.3326 0.2979 10.44 3 0.7205 0.6623 8.08 4 0.4137 0.3828 7.47 5 0.4752 0.4258 10.39 6 0.5703 0.5352 6.15 7 0.5618 0.5501 2.08 8 0.6410 0.5285 17.55 9 0.4630 0.3927 15.18 10 0.4201 0.4032 4.03
下载: 导出CSV
-
[1] FERRER A J, HERNÁNDEZ F, TORDSSON J, et al. OPTIMIS: a holistic approach to cloud service provisioning[J]. Future Generation Computer Systems, 2012, 28(1): 66-77 doi: 10.1016/j.future.2011.05.022 [2] 董元发, 郭钢, 杜轩. 网上网下相结合的云设计服务模式与平台体系架构研究[J]. 现代制造工程, 2016(1): 53-60DONG Y F, GUO G, DU X. Research of online and offline combined cloud design service mode and platform architecture[J]. Modern Manufacturing Engineering, 2016(1): 53-60 (in Chinese) [3] 陈健, 莫蓉, 初建杰, 等. 云设计制造模式下的社交化协作团队构建方法[J]. 浙江大学学报, 2019, 53(3): 444-454 doi: 10.3785/j.issn.1008-973X.2019.03.005CHEN J, MO R, CHU J J, et al. Construction of social collaboration team in cloud design and manufacturing mode[J]. Journal of Zhejiang University , 2019, 53(3): 444-454 (in Chinese) doi: 10.3785/j.issn.1008-973X.2019.03.005 [4] LI X, SONG J D, HUANG B Q. A scientific workflow management system architecture and its scheduling based on cloud service platform for manufacturing big data analytics[J]. The International Journal of Advanced Manufacturing Technology, 2016, 84(1): 119-131 [5] ZHANG L, LUO Y L, TAO F, et al. Cloud manufacturing: a new manufacturing paradigm[J]. Enterprise Information Systems, 2014, 8(2): 167-187 doi: 10.1080/17517575.2012.683812 [6] YAMATO Y, SHIGEMATSU N, MIURA N. Evaluation of agile software develeopment method for carrier cloud service platform development[J]. IEICE Transactions on Information and Systems, 2014, E97-D(11): 2959-2962 doi: 10.1587/transinf.2014EDL8094 [7] 李雪瑞, 余隋怀, 初建杰. 云制造模式下采用Rough-ANP的机械设计知识优选推送策略[J]. 机械科学与技术, 2018, 37(9): 1387-1395LI X R, YU S H, CHU J J. Optimal selection strategy of mechanical design knowledge via Rough-ANP in cloud manufacturing environment[J]. Mechanical Science and Technology for Aerospace Engineering, 2018, 37(9): 1387-1395 (in Chinese) [8] 王有远, 赵璐, 张乐恩. 基于情境约束的知识个性化推送[J]. 中国机械工程, 2017, 28(15): 1812-1819 doi: 10.3969/j.issn.1004-132X.2017.15.008WANG Y Y, ZHAO L, ZHANG L E. Knowledge personalized push based on context constraint[J]. China Mechanical Engineering, 2017, 28(15): 1812-1819 (in Chinese) doi: 10.3969/j.issn.1004-132X.2017.15.008 [9] 那惠珍, 李彦, 熊艳, 等. 面向产品创新设计的个性化知识推送研究[J]. 机械设计与制造, 2016(11): 261-264 doi: 10.3969/j.issn.1001-3997.2016.11.067NA H Z, LI Y, XIONG Y, et al. Research on personalized knowledge push for product innovative design[J]. Machinery Design & Manufacture, 2016(11): 261-264 (in Chinese) doi: 10.3969/j.issn.1001-3997.2016.11.067 [10] AJAO T A, DERIS S. Optimal web service selection with consideration for user's preferences[J]. International Journal of Computer Science Issues, 2013, 10(2): 355-359 [11] XU Y H, YIN G F, NIE Y, et al. Research on an active knowledge push service based on collaborative intent capture[J]. Journal of Network and Computer Applications, 2013, 36(6): 1418-1430 doi: 10.1016/j.jnca.2013.04.010 [12] ZHANG S Y, GU Y, YI G D. A hybrid knowledge push method based on trust-aware and item-cluster oriented to product design[J]. New Generation Computing, 2019, 37(3): 339-357 doi: 10.1007/s00354-019-00053-3 [13] HE W, XU L D. A state-of-the-art survey of cloud manufacturing[J]. International Journal of Computer Integrated Manufacturing, 2015, 28(3): 239-250 doi: 10.1080/0951192X.2013.874595 [14] 初建杰, 李雪瑞, 余隋怀. 面向工业设计全产业链的云服务平台关键技术研究[J]. 机械设计, 2016, 33(11): 125-128CHU J J, LI X R, YU S H. Key technologies of cloud service platform oriented to the whole chain of industrial design[J]. Journal of Machine Design, 2016, 33(11): 125-128 (in Chinese) [15] 樊佳爽, 余隋怀, 初建杰, 等. 工业设计云服务平台下基于用户偏好的设计团队成员优选决策[J]. 计算机集成制造系统, 2019, 25(11): 2863-2873FAN J S, YU S H, CHU J J, et al. Optimal selection method for design team members based on user preference in industrial design cloud service platform[J]. Computer Integrated Manufacturing Systems, 2019, 25(11): 2863-2873 (in Chinese) [16] 任磊, 任明仑. 基于加权协同网络的制造服务组合方法[J]. 机械工程学报, 2018, 54(16): 70-78 doi: 10.3901/JME.2018.16.070REN L, REN M L. Manufacturing service composition method based on service weighted synergy network[J]. Journal of Mechanical Engineering, 2018, 54(16): 70-78 (in Chinese) doi: 10.3901/JME.2018.16.070 [17] 程功勋, 刘丽兰, 林智奇, 等. 面向用户偏好的智能云服务平台研究[J]. 中国机械工程, 2012, 23(11): 1318-1323, 1336 doi: 10.3969/j.issn.1004-132X.2012.11.013CHENG G X, LIU L L, LIN Z Q, et al. Intelligent cloud service platform for customer preference[J]. China Mechanical Engineering, 2012, 23(11): 1318-1323, 1336 (in Chinese) doi: 10.3969/j.issn.1004-132X.2012.11.013 [18] BEZDEK J C. Pattern recognition with fuzzy objective function algorithms[M]. New York: Plenum Press, 1981 [19] NISHA C C, MOHAN A. A social recommender system using deep architecture and network embedding[J]. Applied Intelligence, 2019, 49(5): 1937-1953 doi: 10.1007/s10489-018-1359-z [20] 张芳兰, 陈玉, 李帅. 基于全局HIEs解构的人机产品用户需求模型构建[J]. 图学学报, 2019, 40(2): 303-307ZHANG F L, CHEN Y, LI S. Construction of user need model for ergonomic product based on the global HIEs deconstruction[J]. Journal of Graphics, 2019, 40(2): 303-307 (in Chinese) [21] REBALA G, RAVI A, CHURIWALA S. An introduction to machine learning[M]. Switzerland: Springer, 2019 -
点击查看大图
图(3) / 表(7)
计量
- 文章访问数: 77
- HTML全文浏览量: 27
- PDF下载量: 4
- 被引次数: 0
