Automating FDLP eXchange Withdrawal Lists

Sarah Billman, James McCready, Heather Ross, and Andrew Dudash

The Penn State University Libraries (University Libraries) is a large academic research library comprising twenty-four locations and thirty campus libraries across the Commonwealth of Pennsylvania. The University Park campus is the main campus and has five libraries with many different departments that often act independently. As an example, in the main Pattee-Paterno Library, there are four libraries: George and Sherry Middlemas Arts and Humanities/Social Sciences/Education, Life Sciences (part of STEM libraries), William and Joan Schreyer Business, the Donald W. Hamer Center for Maps and Geospatial Information (Maps and GIS). As a large institution, the scope of projects can be substantial and create opportunities for departments to share their technical skills. These partnerships often manifest as technological innovations that can improve workflow across the organization and be shared with the greater library community.

The University Libraries serves all residents of Pennsylvania. In that role, it is a selective Federal Depository Library. In 1993, The University Libraries entered a selective housing agreement for maps with the regional Depository Library, the State Library of Pennsylvania. The University Libraries already had a collection of other government-issued maps, plenty of growth space, and the staff and expertise to handle this arrangement. Over the past three decades, the influx of maps from the Federal Depository Library Program (FDLP) has strained certain map collection areas and created a need to withdraw material to make space for future needs.

Even though the number of maps sent annually has dramatically decreased in the last ten years, Maps and GIS continues to receive physical maps as part of the FDLP program. One part of the collection growing beyond available space includes western states like Idaho, Colorado, etc.

While investigating candidates for withdrawal, we determined that we had many “Motor Vehicle Use” maps of national forests, mostly west of the Mississippi River. These maps are often released annually with few changes and are of locations outside Pennsylvania and surrounding states. In consultation with our Regional Depository Librarian, we determined that these maps could be treated as superseded and consequently withdrawn—keeping the latest edition only. The older maps could be offered to other depository libraries using the FDLP eXchange program.

Several issues complicated the project. The University Libraries reclasses all depository maps in the Maps and GIS collection to the Library of Congress (LC) classification system, meaning that these maps were interfiled with existing LC-classed maps. Each map sheet is given an individual barcode. Over the years, these maps were inconsistently assigned call numbers, creating call numbers and subject cutter variations. As student employees assessed the maps for withdrawal, the inconsistency of call numbers made it challenging to determine and verify the correct superseded map.

Once we discovered significant variability in the call numbers and subject headings, we realized we needed to implement some standardization moving forward. The Maps and GIS staff worked with the maps catalogers to set rules for cataloging the maps. In the LC classification system, maps are classed by geographic region, then by subject, and finally, in this case, we added the ranger district name at the end of the call number for further differentiation. For example, the call number for the Tongass National Forest in Alaska went from A 13.28:T 61/26/2020 to G4372.T6E63 2020.U5 Wrangell RD. New depository maps will be cataloged using this method, standardizing the call number using the “E63” subject heading (recreation) and adding the ranger district name at the end of the call number. Standardizing the call number will make it easier for student employees to replace superseded maps with the latest edition accurately. There are no plans to make retroactive changes to the existing map call numbers.

In the initial stage of the withdrawal project, nearly 600 maps were identified as Motor Vehicle Use maps, with 350 of those identified as superseded. Student employees pulled the maps to be withdrawn based on a shelf list generated by the library management system’s reporting module. The University Libraries use BLUEcloud Analytics from SirsiDynix Symphony. The next step was to populate the FDLP eXchange template in Excel, which had eight required fields to populate. The title, classification type, document number, number of pieces, publication start date, shipping date, format, and condition are required. During our first attempt at using the FDLP eXchange site, we manually created the list to populate the FDLP eXchange template by copying and pasting from the library catalog. After some trial and error, we loaded the sheets and waited for matches, and we felt like we learned something new about the process as the maps worked their way through the system.

After this initial process, Maps and GIS staff thought the process would benefit from automation as we felt the workflow of copying and pasting eight fields for each map could have been more sustainable, and several of those fields could have been populated from the MARC record. Not having the skills to automate the process, Maps and GIS staff reached out to the Subject Libraries team, which comprises employees from different departments across the Pattee-Paterno Library. The team had developed a method of combining library management system reports with Excel and a wireless barcode scanner to create a cloud-based physical inventory scanning system that was utilized during a large-scale inventory of the Humanities, Social Sciences, and Education Libraries. Maps and GIS staff learned about this project and requested that it be modified to populate the FDLP eXchange template to reduce data entry. After testing the template and giving feedback on the exact values needed for uploading, the Maps and GIS staff successfully tested the workflow by scanning eighty maps and uploading the file to FDLP eXchange. Some manual entry of condition and postage reimbursement remained but was manageable. One significant unresolved issue was the errors generated from items violating the FLDP rule prohibiting the withdrawal of materials at selective FDLP depository libraries that are less than five years old. In this case, these items could still be withdrawn as they were superseded, but the FDLP eXchange site would mark them as errors and have to be manually accepted on the website. Setting the dates to “no date” is not recommended as that may inhibit items matching in FDLP eXchange. One helpful tip when filling out the template is to use the FDLP eXchange data dictionary (https://www.fdlp.gov/sites/default/files/fdlp-exchange/FDLP%20eXchange%20Data%20Dictionary%20v3.pdf) for exact values. For example, the FDPL eXchange wants “Library of Congress” in the classification column, not “LoC.” Using the “LoC” abbreviation will cause errors.

For more information on this process and a sample Excel sheet, see the Penn State University Libraries Government Information LibGuide (https://guides.libraries.psu.edu/government-information/automated).

The technical process for creating the reports for the scanning tool begins with downloading the template from the FLDP website (https://www.fdlp.gov/instruction/fdlp-exchange). Then a shelf list of all our maps collection was generated from BlueCloud Analytics and placed in the Raw tab of the Excel spreadsheet. The report included Call Number, Barcode or another item identifier, Item Title, Classification scheme, Publication Year (if available), and shipping year (often the same as the Publication Year). Our maps are cataloged at the sheet level, facilitating barcodes as the matching value.

Excel lookup formulas were written for each column of the FLDP eXchange template. We chose to use “INDEX/MATCH” for compatibility and flexibility. In this case, the formula matched the scanned barcode in the FDLP template in tab one with the barcode in the shelf list in the Raw tab. It then populated the desired data from the shelf list appropriate column on the FLDP eXchange template. We wrapped these functions in IF statements for purely aesthetic purposes to keep the cells blank if the dumped barcode cell is blank. It keeps the sheet clean while waiting to be filled with barcodes. Figure 1 is an example of the formula.

After the barcodes were added and the information was populated, we carefully examined the file and made any manual edits needed. We then copied and pasted the values from the Template tab into a fresh FLDP eXchange template. We kept the original file to use again.

There are several limitations and potential issues:

• The lookup formula only searches for exact matches. Non-matches are denoted by “---No record found---” and will need to be added manually
• Newer versions of Excel have more powerful lookup functions. The above formula includes the IFERROR function, which may not work in older versions.
• Mismatched Excel formatting (barcodes formatted as text vs. number).
• Leading zeros (barcodes requiring leading zeros to match correctly)
• When dates were unavailable from the MARC record, we pulled the dates from the call number, requiring additional segments in the formula.

Even though this process is partially automated, it generates serious time savings at many levels. This process also has implications for harvesting metadata from the catalog to import into other platforms and other applications yet to be determined.

Sarah Billman (sjh31@psu.edu), Information Resources and Services Support Specialist 3, Penn State University. James McCready (jwm28@psu.edu), Information Resources and Services Support Specialist 3, Penn State University. Heather Ross (hdr10@psu.edu), Map Specialist, Donald W. Hamer Center for Maps and Geospatial Information, Penn State University. Andrew Dudash (dudash@psu.edu), Librarian for Political Science, Policy Studies, and Government Information, Penn State University Libraries.