Insurance consumer information data lake

保险消费者信息数据湖

• 批准号: 520480-2017
• 负责人: Khoury, Richard
• 金额: $ 0.91万
• 依托单位: Université Laval
• 依托单位国家: 加拿大
• 项目类别: Engage Plus Grants Program
• 财政年份: 2017
• 资助国家: 加拿大
• 起止时间: 2017-01-01 至 2018-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=639621
• 关键词: Insurance; consumer; information; data; lake

The Canadian insurance company La Capitale Assurances Générales has collected over time several massivedatasets about their customers, insurance plans, and various insured items. They are now looking for ways ofefficiently combining these datasets together, enrich it with online data, and inferring new knowledge aboutcustomers from it, with the long-term goal of mining this data to learn more information about their customers'habits and lifestyles, in order to customize their insurance coverage to each individual customer's needs. In theoriginal Engage grant, we designed learning algorithms to discover this knowledge from their data, andimplemented a simple database to centralize the information into one source. In this follow-up Engage Plusgrant, we will explore more advanced methods focusing on resolving the issue of the imbalanced class data intheir dataset, an issue that was discovered during the work on the Engage grant, and we will improve thedatabase system. The insurance industry is a multi-billion dollar industry, and the data management tools andknowledge discovery algorithms we will develop for La Capitale Assurances Générales will give Canadiancompanies an important edge in this competitive international marketplace.

加拿大保险公司La Capitale Assumption Générales收集了关于其客户，保险计划和各种保险项目的多个数据集。他们现在正在寻找有效地将这些数据集结合在一起的方法，用在线数据丰富它，并从中推断出有关客户的新知识，长期目标是挖掘这些数据，以了解有关客户习惯和生活方式的更多信息，以便根据每个客户的需求定制他们的保险范围。在最初的Engage资助中，我们设计了学习算法来从他们的数据中发现这些知识，并实现了一个简单的数据库来将信息集中到一个源中。在后续的Engage Plusgrant中，我们将探索更先进的方法，专注于解决数据集中的不平衡类数据问题，这是在Engage grant工作期间发现的问题，我们将改进数据库系统。保险业是一个价值数十亿美元的行业，我们将为La Capitale Assumption Générales开发的数据管理工具和知识发现算法将使西班牙公司在这个竞争激烈的国际市场中获得重要优势。