面向电商领域的旗袍知识图谱构建与实证研究

Construction and empirical research of cheongsam knowledge graph for e⁃commerce domain

为挖掘旗袍服饰要素的关联关系与隐藏价值,构建一个基于电商数据的细粒度旗袍知识图谱,并开展实证研究。旗袍知识图谱采用自底向上逐层构建的方法,首先根据先验知识设计细粒度本体模型,为后续工作提供逻辑基础;然后采集并预处理原始数据,通过知识抽取、知识融合规范化数据,进行实例填充,完成数据层设计;最后将数据存储到图数据库,实现可视化。在此基础上,围绕旗袍、服饰要素以及用户人群开展深度挖掘、知识推理,借助图分析算法在社区检测、相似度和要素关联上展开实证研究。结果表明:旗袍知识图谱蕴含服饰要素之间隐藏的关系,能够挖掘内在的语义信息;旗袍服饰要素和用户人群之间存在一定的知识关联性,进而辅助设计与销售决策。

In order to explore the association relationship and hidden value of cheongsam clothing elements,a fine-grained cheongsam knowledge graph based on e-commerce data is constructed and empirical research is carried out.The cheongsam knowledge graph is constructed layer by layer from the bottom up.A fine-grained ontology model is designed based on a priori knowledge to provide a logical basis for the subsequent work.The raw data is collected and pre-processed,and they are normalised and filled by means of knowledge extraction and knowledge fusion to complete the design of the data layer.The data are stored in the graph database to achieve the visualisation.On this basis,deep mining and knowledge reasoning are carried out around cheongsam,clothing elements and consumer populations,and the empirical research is carried out with the help of graph analysis algorithms on social detection,similarity and element association.The results show that the cheongsam knowledge graph contains hidden relationships between clothing elements,and is capable of mining the intrinsic semantic information;there is a certain degree of knowledge correlation between the dress elements of cheongsam and the consumer population,which can assist in design and sales decisions.