公开(公告)号：US08265984B2

公开(公告)日：2012-09-11

申请号：US12117540

申请日：2008-05-08

申请人： Xin Xin Bai ,  Jin Dong ,  Ta-Hsin Li ,  Ming Xie ,  Wen Jun Yin ,  Bin Zhang ,  Cindy Q. Zhang

发明人： Xin Xin Bai ,  Jin Dong ,  Ta-Hsin Li ,  Ming Xie ,  Wen Jun Yin ,  Bin Zhang ,  Cindy Q. Zhang

IPC分类号： G06Q30/00

CPC分类号： G06Q10/04 ,  G06Q10/0637

摘要： A method and system for integrating multiple factors into a unified optimization model for retail network configuration, in one aspect, obtains input data for modeling store configuration. The input data may include demand of each merchandise category from each customer segment in each facility, geographic distribution of stores in an area, current revenue of stores, and physical cost of reconfiguring stores. A trade area is generated as a function of store location, store format, and store capacity. The method and system also generates trade area demand summation representing predicted total demand of all stores for all merchandise categories for all customer segments in the trade area, as a function of store location, store format, store capacity, merchandise category, and customer segment associated with the trade area. An objective function is constructed as a function of said trade area demand summation, current revenue of stores, and physical cost of reconfiguring stores.

摘要翻译： 一种将多个因素整合到零售网络配置的统一优化模型中的方法和系统，一方面获取用于建模商店配置的输入数据。 输入数据可以包括每个设施中的每个客户段的每个商品类别的需求，区域中的商店的地理分布，商店的当前收入以及重新配置商店的物理成本。 作为商店位置，商店格式和商店容量的函数生成交易区域。 该方法和系统还根据商店位置，商店格式，商店容量，商品类别和相关客户细分，生成代表业务区域中所有客户区域的所有商品类别的所有商店的预期总需求的交易区域需求总和 与贸易区。 构建目标函数作为所述贸易区域需求求和，商店的当前收入以及重新配置商店的实际成本的函数。

公开(公告)号：US20110040601A1

公开(公告)日：2011-02-17

申请号：US12539256

申请日：2009-08-11

申请人： Xin Xin Bai ,  Jin Dong ,  Ta-Hsin Li ,  Wen Jun Yin ,  Bin Zhang

发明人： Xin Xin Bai ,  Jin Dong ,  Ta-Hsin Li ,  Wen Jun Yin ,  Bin Zhang

IPC分类号： G06Q10/00

CPC分类号： G06Q30/02 ,  G06Q30/0201 ,  G06Q30/0204

摘要： A method and system for customer segmentation using adaptive spectral clustering may include determining initial segmentation labels, determining new customer behavior data, formulating a single objective minimization function that integrates the initial segmentation labels with the new customer behavior data, and determining best fit to both the initial segmentation labels and the new customer behavior data simultaneously by minimizing the single objective minimization function.

摘要翻译： 用于使用自适应光谱聚类的客户分割的方法和系统可以包括确定初始分割标签，确定新的客户行为数据，制定将初始分割标签与新的客户行为数据集成的单个目标最小化函数，以及确定最佳匹配 通过最小化单个目标最小化函数同时初始分割标签和新的客户行为数据。

公开(公告)号：US20090281869A1

公开(公告)日：2009-11-12

申请号：US12117540

申请日：2008-05-08

申请人： Xin Xin Bai ,  Jin Dong ,  Ta-Hsin Li ,  Ming Xie ,  Wen Jun Yin ,  Bin Zhang ,  Cindy Q. Zhang

发明人： Xin Xin Bai ,  Jin Dong ,  Ta-Hsin Li ,  Ming Xie ,  Wen Jun Yin ,  Bin Zhang ,  Cindy Q. Zhang

IPC分类号： G06Q10/00

CPC分类号： G06Q10/04 ,  G06Q10/0637

摘要： A method and system for integrating multiple factors into a unified optimization model for retail network configuration, in one aspect, obtains input data for modeling store configuration. The input data may include demand of each merchandise category from each customer segment in each facility, geographic distribution of stores in an area, current revenue of stores, and physical cost of reconfiguring stores. A trade area is generated as a function of store location, store format, and store capacity. The method and system also generates trade area demand summation representing predicted total demand of all stores for all merchandise categories for all customer segments in the trade area, as a function of store location, store format, store capacity, merchandise category, and customer segment associated with the trade area. An objective function is constructed as a function of said trade area demand summation, current revenue of stores, and physical cost of reconfiguring stores.

摘要翻译： 一种将多个因素整合到零售网络配置的统一优化模型中的方法和系统，一方面获取用于建模商店配置的输入数据。 输入数据可以包括每个设施中的每个客户段的每个商品类别的需求，区域中的商店的地理分布，商店的当前收入以及重新配置商店的物理成本。 作为商店位置，商店格式和商店容量的函数生成交易区域。 该方法和系统还根据商店位置，商店格式，商店容量，商品类别和相关客户细分，生成代表业务区域中所有客户区域的所有商品类别的所有商店的预期总需求的交易区域需求总和 与贸易区。 构建目标函数作为所述贸易区域需求求和，商店的当前收入以及重新配置商店的实际成本的函数。

公开(公告)号：US20080046864A1

公开(公告)日：2008-02-21

申请号：US11765001

申请日：2007-06-19

申请人： Xin Xin Bai ,  Ying Liu ,  Ying Hua Liu ,  Jian Wang ,  Xin Zhou ,  Jun Zhu

发明人： Xin Xin Bai ,  Ying Liu ,  Ying Hua Liu ,  Jian Wang ,  Xin Zhou ,  Jun Zhu

IPC分类号： G06F9/44

CPC分类号： G06F8/74 ,  G06F8/35 ,  G06F8/70

摘要： Initial source and target models are divided into segments responsive to at least one segmentation rule. After changes, post-change source and target models are likewise divided. The at least one segmentation rule is defined responsive to at least one transformation rule such that use of the at least one segmentation rule divides initial source and target models into corresponding segments. Change statuses are identified regarding initial segments relative to post-change segments of the respective models responsive to comparing initial segments to post-change segments of the source model and initial segments to post-change segments of the target model. An indication of conflicts between the post-change source model and post-change target model is generated for presentation to a user or to a computer automated conflict settlement process. The generating is responsive to comparing the identified change statuses of the corresponding segments of the initial source model and initial target model.

摘要翻译： 初始源和目标模型根据至少一个分割规则被分成段。 变更后，变更后的来源和目标模型也同样划分。 响应于至少一个变换规则来定义至少一个分割规则，使得至少一个分割规则的使用将初始源和目标模型分成相应的段。 响应于将源模型和初始段的后变化段与初始段相比较以改变目标模型的段，确定关于相对于相应模型的后变化段的初始段的变化状态。 产生改变后的源模型和变更后目标模型之间的冲突的指示，以呈现给用户或计算机自动冲突解决过程。 生成器响应于比较初始源模型和初始目标模型的相应段的所识别的改变状态。

公开(公告)号：US08056049B2

公开(公告)日：2011-11-08

申请号：US11765001

申请日：2007-06-19

申请人： Xin Xin Bai ,  Ying Liu ,  Ying Hua Liu ,  Jian Wang ,  Xin Zhou ,  Jun Zhu

发明人： Xin Xin Bai ,  Ying Liu ,  Ying Hua Liu ,  Jian Wang ,  Xin Zhou ,  Jun Zhu

IPC分类号： G06F9/44 ,  G05B13/02

CPC分类号： G06F8/74 ,  G06F8/35 ,  G06F8/70

摘要： Initial source and target models are divided into segments responsive to at least one segmentation rule. After changes, post-change source and target models are likewise divided. The at least one segmentation rule is defined responsive to at least one transformation rule such that use of the at least one segmentation rule divides initial source and target models into corresponding segments. Change statuses are identified regarding initial segments relative to post-change segments of the respective models responsive to comparing initial segments to post-change segments of the source model and initial segments to post-change segments of the target model. An indication of conflicts between the post-change source model and post-change target model is generated for presentation to a user or to a computer automated conflict settlement process. The generating is responsive to comparing the identified change statuses of the corresponding segments of the initial source model and initial target model.

摘要翻译： 初始源和目标模型根据至少一个分割规则被分成段。 变更后，变更后的来源和目标模型也同样划分。 响应于至少一个变换规则来定义至少一个分割规则，使得至少一个分割规则的使用将初始源和目标模型分成相应的段。 响应于将源模型和初始段的后变化段与初始段相比较以改变目标模型的段，确定关于相对于相应模型的后变化段的初始段的变化状态。 产生改变后的源模型和变更后目标模型之间的冲突的指示，以呈现给用户或计算机自动冲突解决过程。 生成器响应于比较初始源模型和初始目标模型的相应段的所识别的改变状态。