融合农村居民意图的健康知识推荐方法  

Health knowledge recommendations for rural residents incorporating user intention

作　　者：王馨悦 吴华瑞[2,3,4] 陈雯柏 韩笑[2,3,4] 朱华吉 赵春江[2,3,4] WANG Xin-yue;WU Hua-rui;CHEN Wen-bai;HAN Xiao;ZHU Hua-ji;ZHAO Chun-jiang(School of Automation,Beijing Information Science and Technology University,Beijing 100192,China;Information Technology Research Center,Beijing Academy of Agriculture and Forestry Sciences,Beijing 100097,China;National Engineering Research Center for Information Technology in Agriculture,Beijing 100097,China;Key Laboratory of Digital Village Technology,Ministry of Agriculture and Rural Affairs,Beijing 100097,China)

机构地区：[1]北京信息科技大学自动化学院,北京100192 [2]北京市农林科学院信息技术研究中心,北京100097 [3]国家农业信息化工程技术研究中心,北京100097 [4]农业农村部数字乡村技术重点实验室,北京100097

出　　处：《计算机工程与设计》2024年第7期2142-2149,共8页Computer Engineering and Design

基　　金：科技创新2030—“新一代人工智能”重大基金项目(2021ZD0113604);国家重点研发计划基金项目(2019YFD1101105、2020YFD1100602)。

摘　　要：为提高农村居民健康认知水平,提出一种融合农村地区居民意图的健康知识推荐方法。考虑到村民对不同健康知识的偏好,建模用户层中村民与健康知识间的关系路径,捕获村民获取健康知识的潜在意图,利用图卷积与注意力机制传播聚合村民意图邻居节点信息,获得村民与健康知识的高阶邻域表示,通过双交叉聚合器将初始节点与邻域表示进行聚合,增强村民与健康知识的表示能力,实现村民健康知识精确推荐。基于农村健康知识数据集验证研究模型有效性,结果表明该模型在准确率、NDCG指标上均得到了一定程度的提高。To improve the health awareness of rural residents,a health knowledge recommendation method that incorporated the intentions of rural residents was proposed.Considering villagers’preferences for different relationships of health knowledge,the relations path between villagers and health knowledge was modeled as the potential intentions of villagers for acquiring health knowledge at the user level.The higher-order neighborhood representations of villagers and health knowledge were obtained using graph convolution network and attention mechanism to propagate and aggregate villagers’intentions neighborhood information.To enhance villagers and health knowledge node representation,a double-cross aggregator was selected to aggregate the entity and neighborhood information,to achieve accurate recommendation of villagers’health knowledge.The validity of the research model was verified based on the rural health knowledge data set.The results show that the precision and NDCG indexes of the proposed model are improved to a certain extent.

关 键 词：农村 知识推荐 注意力机制 知识图谱 嵌入传播 健康知识 神经网络 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]