摘要
对电子商务多用户相似货源信息进行检索,能够有效提高电子商务交易效率。对电子商务多用户相似货源信息的检索,需要构建用户的词项子空间,对最大检索词团进行提取,完成多用户相似货源信息的检索。传统方法通过D-S证据理论对不确定性语义特征进行描述,获得信息文本特征和主题特征,但忽略了对最大检索词团的获取,导致检索精度偏低。提出基于Markov网络的电子商务交易下多用户相似货源信息检索方法,通过构建词项子空间和文档子空间的Marlov网络,对最大检索词团进行提取,并通过最大检索词团信息建立基于迭代方法的多用户相似货源信息检索模型。根据用户所需信息相关概率确定检索词之间的相关性,通过检索词间相关性计算货源信息的相似性,优化货源信息检索模型。实验结果表明,所提方法信息检索速度较快,检索过程较为简单,且用户对检索结果的满意度较高。
To retrieve similar freight source information of e - commerce user can effectively improve the efficiency of e - commerce transaction. The traditional methods ignore the acquisition of the maximum retrieval word group, which results in low retrieval precision. Therefore, a method to retrieve multi - user similar source information based on Markov network was proposed. By constructing Markov network of document subspaee and lexieal item subspace, the maximum retrieval word group was extracted. Meanwhile, the model of multi - user similar source information retrieval based on iterative method was established by the information of the maximum retrieval word group. According to the relevant probability of information needed by users, the correlation between retrieval words was determined. Finally, the correlation between retrieval words was used to calculate the similarity of freight source information and thus to optimize the information retrieval model of freight source. Simulation results show that the proposed method has fast information retrieval and simple retrieval process. Meanwhile, the user has high satisfaction degree with the retrieval result.
作者
谷钰
张丽杰
吕翘楚
GU Yu, ZHANG Li - jie, LV Qiao - chu(Changchun University of Technolog, Changchuan Jilin 130000, China)
出处
《计算机仿真》
北大核心
2018年第10期472-476,共5页
Computer Simulation
基金
互联网+智慧居家养老信息服务平台(20170505MA2)
关键词
电子商务交易
多用户
货源信息
检索方法
E- commerce transaction
Multiuser
Information of freight source
Retrieval method
