Data unification for customer profile generation

用于生成客户档案的数据统一

• 批准号: 543346-2019
• 负责人: Antonie, Luiza
• 金额: $ 1.82万
• 依托单位: University of Guelph
• 依托单位国家: 加拿大
• 项目类别: Engage Grants Program
• 财政年份: 2019
• 资助国家: 加拿大
• 起止时间: 2019-01-01 至 2020-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=680421
• 关键词: Data; unification; customer; profile; generation

Data unification (i.e., unifying data from different data sources of customers, through intelligent cleaning and matching algorithms to generate customer profiles) is paramount in generating customer data profiles. However, data unification is a complex and challenging problem when working with real-life data. The main challenge is that unique identifiers do not exist, thus one must use common characteristics to all of the databases and compare their values to determine similarity. Other challenges are presented by different database formats, typographical errors and missing data. Finally, processing large amounts of records (i.e., billions) when integrating multiple data sources presents significant computation challenges. Creating customer profiles can be achieved by applying advanced artificial intelligence and machine learning techniques that can be employed to identify and link customers across different data sources. In this project, we will investigate and develop a new custom intelligent system to generate customer data profiles.

数据统一(即，通过智能清理和匹配算法统一来自不同客户数据源的数据，以生成客户简档)在生成客户数据简档时至关重要。然而，在处理真实数据时，数据统一是一个复杂且具有挑战性的问题。主要的挑战是不存在唯一的识别符，因此必须对所有数据库使用共同特征，并比较它们的值以确定相似性。其他挑战还包括不同的数据库格式、排版错误和数据丢失。最后，在集成多个数据源时处理大量(即数十亿)记录带来了巨大的计算挑战。创建客户配置文件可以通过应用先进的人工智能和机器学习技术来实现，这些技术可用于识别不同数据源的客户并将其链接起来。在这个项目中，我们将研究和开发一个新的定制智能系统来生成客户数据配置文件。