Tagging guide needed!

[cvlowe]

I tried google, but ended up with a lot of "how to apply thousands of tags to your book to increase sales" type of links..

Anyways, I'm just curious how some long-term people use tags.

There are obvious tags like 'fiction', 'science fiction', 'fantasy'. Those are usually pretty easy to figure out and add as necessary (though I'm finding a LOT of fantasy books which have the tag science fiction pre-applied even though there is no "science" in them..)

A couple really bed examples: I imported a couple cookbooks and the tag list seems to include EVERY SINGLE item used as a recipe ingredient. A bit of overkill, then again it might be useful- if I want to make something using 'avocado' I can just look for that tag.. yet having too many tags will limit the usefulness also.

What is the purpose of 'general'? That seems to be on EVERY SINGLE BOOK I've added so far. I've been deleting it from the tag list regularly, but when I aqdd more books, it shows up again.

How detailed should I get? For example- A Spell for Chameleon: Xanth book 1. That obviously gets the tags: fiction, fantasy, Xanth. Should I add tags for other things that appear in the book? Dragons, swords, magic, zombies, centaurs, harpies, etc.. Should I include "unique" stuff like character names?

I realize everybody probably has their own method. I'm simply seeking opinions to help figure out how detailed I want to make my tag method. My biggest hurdle is just trying to figure out exactly how detailed I want to get.

[eschwartz]

I would never add a tag that can be implied by either reading the title or series.

I definitely would not mark books with series:"=Xanth" as tags:"=Xanth", it is nearly as bad as tagging Centaur Aisle "Centaur Aisle".

What I usually do is delete all tags no matter what, then re-add any tags later, when I decide I need it in order to find books of a similar type.

Character names are somewhat of a pet peeve of mine. They serve no useful purpose, because if you are looking for books about a specific character, you already know what book or series of books you are looking for, and can just search based on that. But every metadata source (I don't use them anyway, but it is the theory of the thing) includes them.

[fidvo]

I agree with eschwartz. I never go by other people's tags. The only metadata I download is cover and summary; the rest I fill in by hand.

My tags are hierarchical, but I limit them to two levels. I have, for instance, "Genre.Science Fiction" and "Genre.Mystery" as well as things like "Content.Prehistoric" and "Content.Alien Invasion."

That applies to my Fiction library. My "Special" library has a different set of tags, but again they're limited to just a few levels. For my cookbooks, for instance, I have "Cookbook.Asian.Chinese" and "Cookbook.Mexican." Beyond that, I have just a few basic tags for cookbooks, such as "Cookbook.Healthy," "Cookbook.Frugal," and "Cookbook.Easy" to tag cookbooks that make a claim of being healthy, cheap, or easy to fix. If you keep your cookbooks in a separate library entirely, you could even get rid of the "Cookbook" prefix.

I also have tags such as "Mythology.Monsters" and "How-To.Magic Tricks." Generally, the first part of the tag indicates the broad category while the second part refers to the subject matter.

My one concession to maintaining long lists of entries in a single field is that I have a custom "short stories" field in my Fiction library that contains all the stories in each anthology or collection. That allows me to look up a short story by title.

[BetterRed]

I mainly have non-fiction books, most books only have one tag - eg "US Civil War", "East Asian Food", "Investment Banking" etc

If I want to find all the books that cover the Battle of Antietam, or I want a recipe for Kang-kong then I use Windows Full Text Search. The response time on my ~45,000 book library is usually sub 2s.

For the Kang-kong recipe I would probably flick through the books directly from the search results and copy the chosen recipe to my Galaxy Note and head for the kitchen.

For the Antietam example I would probably use a process that utilises the Import List PI to create a Reading List or I might just save the Windows Search results file into the Library (a recently realised 'light bulb moment' alternative)

The problem with extensive tagging of non-fiction is that it assumes a) you have read the book and b) you already know what you're going to be looking for in the future.

BR

[Inferno]

I think it is more to do with what you will be looking for when you search the tags than what other people do.

For instance, although fantasy is a subset of science fiction historically, it has grown to become huge. forty years ago it was not so booming a genre.

Although my tagging doesn't use hierarchical tags if I was starting today I probably would have done that. I tend to put fiction or non fiction in as tags although many find that unnecessary.

I started using a separate library to do an initial "cleanup" of imported books when adding many so I wouldn't be adding all the extra tags that seem to be as useful as the metadata some websites add lol. Everything plus the kitchen sink.

[DaltonST]

-------

@cvlowe,

Sounds like you need an industrial-strength, table-driven, flexible Tag Scrubber. Download the User Instructions for the Quarantine And Scrub add-on/plug-in from https://www.mobileread.com/forums/sho...d.php?t=247425 , and read about its very powerful Tag Scrubber.

I had the same problem you have, which I why I developed Quarantine And Scrub. Killing off thousands of undesirable Tags forever is trivial if you have Q&S to do it for you.



DaltonST