Fixing Yesterday's Solutions: Data Cleanup in Serials Solutions 360 Core.

Managing electronic serials is a vital part of modern academic librarianship. In 2014, the College of Staten Island (CSI) faced a serious serials access problem. A number of subscribed titles were temporarily not accessible through library interfaces. To quickly solve the problem, a set of over 5,000 custom records were uploaded (using Microsoft Excel) to Serials Solutions 360 Core, CSI's serial management system. This solution provided access to current and back-file (perpetual ownership) titles but introduced a number of significant new problems, including duplicate records, records with holdings dates that were invalid at the end of the calendar year, back-file records that were difficult to identify or edit, and unmanageable sets of records for large subscription packages. This article will describe how and why the custom records caused these issues and describe the project that addressed them. The project (ongoing) was designed to fix the problems while maintaining access to all current and back-file titles during the process. A variety of techniques were required. Vendor-provided knowledgebase records were examined, activated (or deactivated), and downloaded. The Overlap Analysis tool of 360 Core was employed to match knowledgebase and custom (local) records. The custom records were manipulated in Excel for examination, editing, and batch deletion. Excel formulas were used to compare the custom records with vendor knowledgebase records and Overlap Analysis results. Finally, extensive manual work was done to delete records, test access links, and find unmatched records. These techniques have been successful. The project also reaped other unexpected access improvements. Other libraries may apply these techniques and strategies to any serial data clean-up. Although journal management technology is rapidly changing, libraries will still need to be able to work with and manipulate data to provide access, understand and manage their holdings, and audit error-prone vendor data. [ABSTRACT FROM AUTHOR]