Archive for the ‘common knowledge’ Category

Wednesday, June 3rd, 2020

Series Gets a Revamp

series_screenshot

Short Version

Today we roll out a new version of “Series” and “Publisher Series.” Here are some pages to check out:

We’re going to be discussing New Series starting from this Talk post.

The rest of this blog post explains the whys and wherefores in great detail.

“Old” Series

Before today, series were based on the Common Knowledge system. Common Knowledge is a simple “fielded wiki,” a system for keeping and tracking simple values.(1) To add a series to a work, you’d go to the common area of a work page and fill it out as follows:

bryson

It got complex quickly. Here’s one Star Wars book, with stuff inside parentheses for sorting and labeling.

starwars

Needless to say, an entry like “Star Wars (0.0112994350|88.5-22 BBY)” was inaccessible to many. Nor could works be added to a series on the actual series page. Series didn’t extend well to other languages—unless the names coincided, there was endless duplication of effort. A lack of any sort of grouping or subseries gummed up major series with edge-cases, like the re-segmentation of the Lord of the Rings applicable to only some Japanese editions, and made it tricky for users to look at a series and figure out what to read. And while some information came to adhere to series, the whole system was jerry-rigged. Finally, adding NEW features was truly impossible!

It is testimony to the passion and diligence of LibraryThing members over the last 13 years that they have added some 125,000 “regular” series and 30,000 “publisher” series!

“New Series”

New series starts with a more sophisticated data structure and user interface. Series exist as their own, complex entity, like works and authors are, not as series of Common Knowledge “strings.” This means:

  • Adding to series can be done on either work pages or series pages. (On work pages, series have been moved to the (renamed) “Series and work relationships” section.)
  • Sorting works within series is accomplished by dragging and dropping, or by giving the series a default sort, such as by publication or title.
  • Adding labels like “book one” can be done directly, not as part of a larger formula.

Series can now include “groups.” Every series has a “core” grouping, but can also include sections for omnibus editions, short stories, or anything else that—while useful—might be worthwhile to separate out. You can see this on the Lord of the Rings page.

The more sophisticated structure allows for other innovations:

  • A single series can serve across all of LibraryThing’s languages, with different names in different languages.(2)
  • Series can be combined and, in combining, the editor can choose which elements to bring over from one series to another.
  • Series can now be “related” to each other, much as works can be related to works. For example, the Harry Potter Movies can be listed as an adaptation of the famous novels.
  • Every series-related action is separately tracked for examination by members and staff—much like Common Knowledge but with all the extra detail available once single strings were abandoned.

“New Series” has also advanced LibraryThing’s “LT2” redesign project. In making the new pages, Chris Holland essentially worked out LT2 code and concepts, and applied them to a single page on “LT1.” He has learned a lot about how to recast LibraryThing pages without breaking everything.

Finally, series can now be touchstoned, just like authors and works! As works use single brackets, like [War and Peace], and authors use double-brackets, like [[J. K. Rowling]], series use three brackets like [[[Twilight Saga]]].

Future Plans

The near future will see:

  • Members able to follow a series, and see and receive updates when new books are released in that series.
  • “Publisher series” transformed by allowing these work-based lists to be narrowed down to the publishers and editions that pertain to them.

Can You Help?

Series needs your help! Old data needs cleaning up, and all sorts of new data needs adding.

  • We need your help finding bugs and improving existing features so they are maximally intuitive and useful.
  • We need help establishing best practices and norms for the new possibilities. For example, now that we have true series “relationships,” I favor removing adaptations from series and making them and their own series.
  • The biggest data problem is a surfeit of non-English variants. The Common Knowledge structure hid them, but members using LibraryThings other language sites, like LibraryThing.fr (French) and cat.LibraryThing.com (Catalan), created an enormous number of series too—most of them the same as the English series. They need to be combined. For example, before I combined them, the Twilight Saga also existed as “Houkutus” (Finnish), “Saga ‘Zmierzch'” (Polish), and “Crepúsculo” (Spanish).
  • The second biggest task is reviewing the “groups” within series. Omnibus editions and selections have been automatically assigned to a separate group with 95% accuracy, but other groupings have not been attempted.
  • There is a “Needs Help” / “Looks Good” control within the Edit dropdown menu. You can use this to flag the series as needing help or give approval that the series is currently in good shape.

Check It Out

Here are some links to check out!

Here are some links of interest to people who want to dig deeper:


Footnotes:

1. For more on Common Knowledge see our 2007 blog post.

2. Separate series should only be maintained if there is a difference between the series so great that combining them would mislead. This is one of those things we’ll have to hash out as a community.

Labels: common knowledge, new features, series

Thursday, March 21st, 2013

Common Knowledge at 5 million!

Earlier in March, LibraryThing Common Knowledge hit five million edits (and flew right on by: another 53,000 have already been added since then!).

The five-millionth CK fact was … drum roll please … the term “American Revolution” added in the “Important events” field on Barbara Tuchman’s book The First Salute by member berry25. Hey berry25, want an LT t-shirt or a CueCat?

Common Knowledge? What’s that?

Common Knowledge, a part of LibraryThing since 2007, is our vast fielded wiki system of bookish data, capturing everything from characters (Frodo Baggins, C-3PO) to series and awards information to related movies, dedications, author information, and much, much more. See the wiki page for a full rundown.

Some of these pages are ridiculously, awesomely complex: check out the Star Wars series page, for example (872 works, with something like 70 sub-series!). To get a sense of the depth and breadth of everything included in Common Knowledge, check out the clouds page.

Who’s added all this info?

CK would not be the amazing resource that it is without the hard work of the many LibraryThing members responsible for those 5 million+ edits. (Back in 2007 Tim predicted it would prove “insanely addictive”, and that seems to have been spot-on). More than 1,000 LTers have contributed at least 600 edits, and some of the totals are extremely impressive. Here are the top five all-time CK contributors:

Can I contribute?

Please do! It’s super easy to add Common Knowledge data – you’ll see the fields at the bottom of every work or author page on LibraryThing. And if you have questions, thoughts, or suggestions, chime in over at the Common Knowledge, WikiThing, HelpThing group.

Here’s to you all, and here’s to five million more Common Knowledge contributions!

Labels: common knowledge, milestones

Wednesday, July 20th, 2011

Legacy Libraries 2.0: lists, clouds, and more!

Thanks to some fantastic work by Chris Holland (conceptdawg) we’ve just launched a brand new homepage for the Legacy Libraries project, chock full of interesting features and data:

http://www.librarything.com/legacylibraries

It includes the ability to search the contents of Legacy Libraries (LLs) as a whole or by selected subsets; you can also browse LLs by category (like Authors or Signers of the Declaration of Independence), and see a whole series of clouds about the libraries.

For each category of Legacy Library, like Authors, we’ve added new status markers (complete, in progress, proposed, unitemized), and you can sort each list by status, name, date, or library size.

We’ve also integrated data about the Legacy Libraries into a slightly modified version of Common Knowledge, so each library, regardless of completion status, now has an LLCK profile (here’s John Adams’) containing data about the person and their library (largely for cloud-creation purposes, among other things). Feel free to augment this data, but please do read the help page first, since there are some differences between this and the way other CK edits are done. Any questions, just let me know (jeremy@librarything.com, or jbd1 on LT).

This LLCK data allows us to do some really interesting things, like display proposed and unitemized libraries well for the first time (example) and also keep better track of project status. We also, at long last, have a way to highlight the many members of LT who’ve worked so hard on these projects over the (nearly) four years we’ve been cataloging Legacy Libraries (see the contributors cloud at the bottom of the page).

You’ll also notice some integration of these new features on profile and author pages, and Chris has whipped up a handy “Featured Legacy Libraries” module for your homepage (by default at the bottom of the right column).

For more on this, see the Talk thread, and as always, let me know if you have data on a library we should add or further information about any one already on our radar. Submissions of library data are always welcomed and appreciated!

Labels: common knowledge, legacies, legacy libraries

Friday, January 21st, 2011

Separate pages for divided authors!

I’ve introduced separate pages for divided authors. It’s very rough so far. Read about it on Talk.

Labels: common knowledge

Tuesday, December 1st, 2009

Related Movies in Common Knowledge

Chris has added a feature for related movies in Common Knowledge. Here’s an example, from Romeo and Juliet.

The edit box “autocompletes” with suggestions by polling the IMDB API. Here’s an example from Room with a View.

Come talk about it. We’re still hashing out the best way to do it.

Labels: common knowledge, movies