Standards Column

In this column we will review the status of the ISO project to create a family of documents related to customer satisfaction. We will discuss a new project on people aspects of quality management. We will review new projects to revise, expand, and update standards for auditing management systems. We will review the status of the ISO series on capability and performance measurement. And we will review a standard on detection and analysis of outliers.     

非对称信息下考虑顾客抱怨的供应链质量控制研究

本文研究存在顾客抱怨的情形下,由一个制造商和一个零售商组成的供应链质量控制问题。构建了制造商和零售商的博弈模型,分析了非对称信息情况下当顾客抱怨和外部故障成本分担系数发生变化时,制造商和零售商的最优质量检测水平的变化,并得出非对称信息情况下制造商的最优质量检测水平随顾客抱怨的增加而线性增加的结论。     

数据仓库和OLAP技术在CRM投诉分析中的应用研究

借助商业智能解决方案技术,充分利用各种网络业务运行信息,通过科学的主动监控技术发现网络业务运行中的潜在问题,实现有效的主动监控预警分析,及时的做出预先警告,并进一步推进集中监控的精细化,提高网络业务运行的安全性、稳定性。重点要以客户投诉数据建立数据仓库,针对不同的分析主体设计建立各自的多维数据集,利用联机分析技术进行多维度分析,得出客观的,值得运营商借鉴的实时动态结果,为电信运营商设备的主动监管提供有力的技术支持,为广大的网络用户提供更优质的服务。     

燃气企业服务投诉与危机管理分析与对策

近年来,深圳燃气一直致力于打造具有行业特色的服务体系,其核心目的就是打造专业、优质的服务能力,创建具有行业特色的公用服务品牌。从而赢得社会、公众的认可,以及赢得客户的信赖,为拓展燃气销售市场提供强有力的保证。而就服务过程来说,不可避免地会遭遇投诉乃至危机,根据"20/80"原理,占20%比例的投诉和危机是服务体系预防的重点。本文试就燃气企业服务投诉和危机管理展开分析,并提出相应的对策。     

Judging the credibility of historic child sexual abuse complainants: How judges describe their decisions.

Criminal prosecutions of child sexual abuse alleged to have occurred in the distant past raise myriad challenges. One significant challenge involves assessing the credibility of complainants. In the present study, 51 bench trials involving 87 complainants were coded into categories related to complainants’ memory for the offense, as well as credibility of the complainant, reliability of the evidence, and judicial inferences. A total of 4,827 judicial comments were identified and categorized. Judges were more likely to describe the allegations specifically than generally; however, they were sensitive to the predictably impoverished nature of memory after such a long delay. Consistent with psychological evidence, there were more judicial comments about inconsistencies in acquit cases than in convict cases and more comments about corroboration in convict than in acquit cases, although neither inconsistencies nor corroboration were strongly associated with verdict. Of some concern was the apparent and considerable judicial interest in complainants’ behavior and emotions around the time of the alleged abuse and around the time of disclosure. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

面向电信客户投诉和建议的智能分析模型

文中提出了一个面向电信客户投诉和建议的自动快速分析模型,通过网络爬虫采集互联网上客户的投诉和建议,通过客服录入获取客户的电话投诉和建议信息,然后对这些投诉和建议文本进行预处理,包括中文分词、消去停用词、构建向量空间模型等。最后使用结合了人工信息和统计信息的自动识别算法对客户的投诉和建议进行分析、识别,把它们转交给相应的部门处理。     

不完美检验和顾客抱怨的供应链质量控制模型

由于最终产品的质量水平由供应商生产质量水平和分销商检验水平共同决定,将来料检验不完美和顾客抱怨纳入供应链质量控制模型。结果得到了双方博弈的均衡解,阐明了解的内涵和解的特征以及参数间的关系,并给出了均衡解为混合策略时解与分销商检验水平和顾客抱怨之间的互动机制。在理论上给出了分销商来料检验不完美情形下的分析模式和解的结构,并在应用上指明了来料检验水平和顾客抱怨对供应链质量控制的调节方法。     

Report of the Ethics Committee, 2001.

Reports to the membership regarding the number and types of ethical complaints that are investigated in accordance with the bylaws of the American Psychological Association. In accordance with the bylaws of the American Psychological Association (APA), the Ethics Committee reports regularly to the membership regarding the number and types of ethical complaints investigated. In addition to processing cases, in 2001 significant activities included forming and implementing changes to the adjudication process, expanding and developing educative efforts, and continuing work on the revision of the "Ethical Principles of Psychologists and Code of Conduct." (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

电力潜在敏感客户预测的大数据方法应用

随着 95598 业务的不断发展延伸,人工话务强度增大。为了进一步加深对客户隐性特征以及诉求的认识和理解,提升 95598 人工精细化客户服务水平,对投诉倾向等客户服务中的典型应用场景进行了需求细化。基于电力服务工单数据,选取建模关键指标,通过熵权法、主成分分析和决策树等数据挖掘算法,对潜在投诉倾向客户和计划停电敏感客户进行识别,以便有针对性地进行服务资源调度,充分做好应对措施,有效减少投诉压力,提升服务精度。     

基于word2vec和XGBoost相结合的国网95598客服 投诉工单分类

为了解决95598客服投诉工单的整理、归档等问题,其中包括：在人工进行归档的过程中出现的疏忽造成的归档随意问题,即归档准确性问题;人工对投诉工单进行差错点归纳的耗时问题,即效率问题;人工对客服投诉分析深度不足,无法精准快速定位用户诉求热点的问题,即深度问题。本文针对以上三个问题给出解决方案,采用word2vec和XGBoost相结合的方式达到对95598客服投诉工单精准归纳。在文本词向量化的过程中采用word2vec方法,得到单词的文本词向量;利用XGBoost算法对95598客服投诉工单进行分类归档,并且对历史投诉工单的责任部门、专业分类、诉求事件、差错点四个方面进行标注。该模型的分类准确率在83%-91%左右,有较好的的效果。基于工单分类的结果,并设计了相关的投诉类看板,更直观的对数据进行展示。