The librarians have the ability to edit book titles (for instance deleting stray extra text and otherwise cleaning up dodgy data). However the search facility on Natively matches on the original unedited title, not the edited title. This means it produces results that don’t make sense. For example:
-
A search for “Pratchett” in English books: English Books Difficulty List | Natively gives two results, including Love Will Tear Us Apart: (The Stranger Times 3) | L30?? which makes no mention of Pratchett in its Natively listing. (The title on Amazon UK kindle is “Love Will Tear Us Apart: (The Stranger Times 3): Don’t miss this riotous supernatural fantasy novel, perfect for fans of Terry Pratchett!”.)
-
Searching Japanese books for 講談社: Japanese Books Difficulty List | Natively produces a lot of books which have had the publisher name edited out of the title by librarians like 死物語 上 | L37 (original title “死物語 上 (講談社BOX)”), as well as ones which still have the stray text in the title, like 人狼サバイバル 才気煥発! 薔薇園の人狼ゲーム (講談社青い鳥文庫 Eあ 8-17) | L24 .
The search should be searching Natively’s corrected and cleaned up data, because that’s what the user actually sees and because it gets rid of wrong results like the Pratchett example.
(Is this happening because we’re searching the wrong thing in the database, or because we don’t recreate the search index data after librarian edits?)
