TCA Letters to the Editor
Press Release
Title: Comments on Misstatements:Review of LexisNexis Serial Set project
Date: May 2004
Organization: LexisNexis
Letter:
To the Editor of The Charleston Advisor:
We were pleased to see a comparative review of the two digitized Serial Set products in the April 2004 issue of The Charleston Advisor.
While the reviewer went to considerable lengths to provide a comprehensive review, there were several misstatements concerning our product, the LexisNexis® U.S. Serial Set Digital Collection, which we would like to take the opportunity to correct:
- PDFs within our product can be viewed with versions of Adobe Reader 5.0 and higher (the review stated 6.0 was necessary).
- We are releasing content on an ongoing basis, not in 4 phases.
- Users can specify any number of characters to complete a proximity search, though the review stated the “proximity search is limited to within 5, 10, or 25 words.” Although our dropdown boxes provide common operators for novice users (i.e. w/5, w/10, and w/25), users can type proximity connectors (w/n or pre/n) directly into a search box and specify any number of characters between 1 and 255.
- Wildcards and truncation work on all search forms, as long as a minimum of 2 characters are used. (According to the review, “the user can use wildcards and truncations only within LexisNexis Guided Search.”)
Perhaps more importantly, however, we take issue with a fundamental assumption underlying the reviewer’s comparison.
The reviewer limited her comparative product searches to a time period that she was aware had not been completed by LexisNexis, and then penalized our product for not having the same number of hits as the competing product.
We have been very clear about the way we are digitizing the Serial Set. Only documents whose OCR accuracy level is equal to or above 98% are placed online. All other documents are diverted into a separate workflow. We chose this process in the interest of making as many documents available to researchers as quickly as possible.
We take the misconception that our product will remain “incomplete” seriously. Consequently, effective June 2004, we will remove any doubt that our product will include particular documents by loading records for all of the documents that were diverted into a separate workflow.
The bibliographic data for these records—title, author, committee, date, and subjects—is complete, but the pdfs that accompany them are considered “working copies” because the underlying OCR is below 98%. Although the underlying OCR is not completely full-text searchable, the pdfs are of sufficient quality to enable students and researchers to review the replica of the original document. Final versions of all pdfs, accompanied by OCR above 98%, will be uploaded before the end of 2005.
This change in operating procedure will accomplish two things:
- Provide reassurance of our collection’s completeness to librarians and students
- Facilitate research through providing earlier access to pdfs and accompanying bibliographic data
All pdfs that are not final will be clearly flagged within the product. Links to affected pdfs will include the following notation: [Final OCR and image pending], so users will clearly be able to differentiate between final and “working” pdfs.
Although a comprehensive and true side-by-side comparison will only be possible when the competitor has completed their product in 2008, we feel confident that librarians have enough information in hand to fairly evaluate the two products and make a decision for their institution that is right for them.
We thank you for undertaking the important work of evaluating products in the marketplace and hope we can continue to work with you to provide timely access and information.
Sincerely,
Andrew Laas
Academic Product Manager
LexisNexis U.S. Serial Set Digital Collection
