Libraries throughout the world use OCLC’s EZproxy software to manage access to e-resources. When cleaned, processed, visualized and enhanced, these logs paint a valuable picture of a library’s impact on researcher’s lives. The purpose of this paper is to share techniques and procedures for enhancing and de-identifying EZproxy logs using Tableau, a data analytics and visualization software, and Tableau Prep, a tool used for cleaning, combining and shaping data for analysis.,In February 2018, The Ohio State University Libraries established an automated daily process to extract and clean EZproxy log files. The assessment librarian created a series of procedures in Tableau and Tableau Prep to union, parse and enhance these files by adding information such as user major, user status (faculty, graduate or undergraduate) and the title of the requested resource. She last stripped the data set of identifiers and applied best practices for maintaining confidentiality to visualize the data.,The data set is currently 1.5m rows and growing. The visualizations may be filtered by date, user status and user department/major where applicable. Safeguards are in place to limit data presentation when filters might reveal a user’s identity.,Tableau used in concert with Tableau Prep allows an assessment librarian to clean and combine data from various sources. Once procedures for cleaning and combining data sources are established, the data driving visualizations can be set to refresh on a set schedule. This expedites the ability of librarians to derive actionable insights from EZproxy data and to share the library’s positive impact on researcher’s lives.

中文翻译：

---

增强和理解EZ Proxy日志的非程序员指南

全世界的图书馆都使用OCLC的EZproxy软件来管理对电子资源的访问。这些日志经过清洗，处理，可视化和增强后，可以很好地描绘图书馆对研究人员生活的影响。本文的目的是分享使用数据分析和可视化软件Tableau和用于清洁，合并和整形分析数据的工具Tableau Prep增强和取消识别EZproxy日志的技术和过程，2018年2月，俄亥俄州立大学图书馆建立了一个自动化的日常流程来提取和清理EZproxy日志文件。评估馆员在Tableau和Tableau Prep中创建了一系列程序，以通过添加用户专业，用户状态（教职员工，研究生或本科生）以及所请求资源的标题。她最后一次删除了标识符的数据集，并应用了最佳做法来维护机密性以使数据可视化。该数据集目前有150万行，并且还在不断增长。可视化可以按日期，用户状态和用户部门/主要人员（如果适用）进行过滤。当过滤器可能泄露用户身份时，可以采取适当措施来限制数据显示。Tableau与Tableau Prep配合使用，使评估馆员可以清理和合并来自各种来源的数据。一旦建立了清理和合并数据源的过程，就可以将数据驱动可视化设置为按设定的时间表刷新。