Quote:
Originally Posted by theducks
I think you answered part of your own question: That 5%
To address the other 95%
Public meta-data is cr*p.
Way too many tags (especially for readers that use the tags to create categories)
Manual, gives you a chance to Normalize the data before the final save.
Code:
Publisher Example:
Baen
Baen Books
Baen Books Pub.
Baen Pub.
Simon and Schuster (their distribution channel)
Junk, Trash, Cr*p 
Same for Authors (with twists: for who is listed first, LN-FN vs Fn Ln issue)
To steal a line from Burger King (tm):
"Do it your way"
 |
I couldn't agree more! The state of available metadata is horrendous.
Most publishers either aren't aware of the issue, aren't motivated to address the issue, don't have resources to address it, etc. (Google has been terrible with metadata for its scanned works.)
I predict it will take several more years for the field to mature before we see metadata done right. Until then I'll always recommend reviewing metadata manually.
Remember: Garbage In = Garbage Out!
Troy