| 基于网络信息挖掘的港口客户信用风险评价 |
| Risk Assessment of Port Customer Credit Based on Data Mining of Network Information |
| 摘要点击 26 全文点击 0 投稿时间：2023-02-02 修订日期：2024-07-23 |
| 查看/发表评论 下载PDF阅读器 |
| 中文关键词 网络信息挖掘; 置信度修正; 风险评价; 港口客户选择 |
| 英文关键词 network information data mining; confidence adjusting; risk evaluation; port customer selection |
| 基金项目 国家自然科学基金项目（面上项目，重点项目，重大项目） |
| 作者 | 单位 | 邮编 | | 宋云婷 | 东北财经大学 | 116025 | | 党延忠 | 大连理工大学 | 116024 | | 王诺 | 大连海事大学 | | | 赵瑞嘉 | 大连海事大学 | |
|
| 中文摘要 |
| 针对港口经营过程中出现的客户选择风险控制问题, 提出基于外部网络信息数据挖掘的港口客户信用风险评价方法. 通过分析港口客户选择时面临的多维风险信息, 对网络新闻资讯进行数据挖掘并分析其情感倾向与等级; 引入半衰期函数修正网络信息时间维度置信度, 采用贝叶斯平均方法解决各情感倾向等级在融合过程中存在的规模不一致问题; 将网络信息数据挖掘与多准则决策方法相结合, 提出非结构化信息与评价体系相融合的解决思路. 研究结果显示, 通过引入外部网络信息, 能够利用全面实时数据多维度甄别客户间的差异, 帮助港口管理者更好地评估和应对客户信用风险, 进而保障港口业务的稳健发展. |
| 英文摘要 |
| Credit risk assessment methods based on data mining of external network information is proposed to address the risk control issues during the process of selecting customers for ports. The confidence in the temporal dimension of network information is adjusted by introducing a half-life function. The Bayesian averaging method is employed to resolve the issue of inconsistent scales among different sentiment levels during the fusion process. A solution approach that combines unstructured information with assessment systems is proposed, by integrating the data mining of network information with multi-criteria decision-making methods. The results show that the method can more comprehensively identify the reputation and potential risks of customers, and can provide more multidimensional assessment information for ports to choose cooperative customers. |
| 关闭 |
