Introducing Syndetics Unbound

Syndetics Unbound—our new library product, co-developed with ProQuest—is finally out. You can read about it here, and see it in action at the Hartford Public Library.

http://blog.librarything.com/thingology/2016/10/introducing-syndetics-unbound/

This topic is for discussion of Syndetics Unbound. What do you think?

Now that Syndetics Unbound is out, some of our time is going to be turning to LibraryThing development. I'm making another topic for talking about LibraryThing development.

Here's that topic: https://www.librarything.com/topic/237958

Really not a fan of the new tag display. Now we have to read all the tags separately instead of having the most prominent ones pop out at us.

Can you post a screenshot of what you're seeing then? it does appear to bold prominent ones and also a different area.

I can't say I'm a fan of the new tag display either (is there a way to dig down to less popular tags?), though I can dig the increased discoverability of tagmashing (though, again, it seems to be a limited/controlled set ... would be nice to have an option to add your own). Will the Syndetics tag display be coming to LT?

How does the "Add a Tag" option at the top interact with LT tags, if at all?

Does the "Book Profile" info come from LT?

What's the logic behind the tag... boxes? I'm not really sure it's an improvement over tag clouds, except for solving the varying font size issue.

I guess my issue with most tag display mechanisms is that they are so non-linear. There's no good way to browse the tags in an orderly fashion. If you go right to left, they aren't in rows. If you go top to bottom, they aren't in columns. How do you know you haven't missed something? You can't even skim, because the format doesn't let you focus enough to tune out the format and process the information.

I'm a fan of tag lists, personally. All of the information, none of the distractions.

(Edited for syntax.)

>4 omargosh: I don't think that tags added on the OPAC end actually interact with/are added to LT data, but Tim can correct me if I'm wrong.

LT users are, I think, largely responding to the tag display as it were an LT feature. But the context is very different. LT tags in the catalog aren't there for people to drill down deeply into them. They're not fun data for data-minded people, least of all fun data about people's own books. They're mostly jumping-off points for exploration.

The problem with tag clouds here is that, besides looking very dated, they create a false sense of importance. If you do them honestly, for example, most fiction books will have, as their largest tag, "fiction." Other common but deeply un-useful tags will be larger than anything interesting too. And everything will be assembled like a ransom note--a welter of different font sizes, communicating data (the number of members on LT who have used the given tag) that doesn't really mean much in the catalog.

The point here was to provide something more readable. Some element of importance is communicated by color and boldness. But mostly we're saying "here are some easy-to-read tags that we think might be useful to you."

though, again, it seems to be a limited/controlled set ... would be nice to have an option to add your own

Yeah, they're terms we've selected as being useful to sort by. 2-3 are simply the most popular. The rest must split the results. That is, it disfavors filtering tags that don't change the result much. It's looking for things that cut it in two.

As for the ability to add tags, well, maybe. But context, again, is important. Library-catalog users are not there to play with the system the way many LT members are. Their motivations are different.

How does the "Add a Tag" option at the top interact with LT tags, if at all?

Not only does it not interact with LT, it's not actually part of Syndetics Unbound. It's part of the catalog system, Encore. It has tagging; nobody uses it. I rather wish Hartford would remove it, for the reasons your confusion illustrates.

Does the "Book Profile" info come from LT?

No, that's from professional classifiers at Bowker/ProQuest/Syndetics.

>7 timspalding: I thought you had said you might bring some of these UI elements into LT. If you're not considering that for the tag view, then I have no concerns.

I think it all looks very cool. Wish my library had it.

>7 timspalding: From an Accessibility perspective, the new take on tags is much better than a cloud, which I'm guessing is very important at a library level.

>7 timspalding: I still say if readability is your biggest goal, a list is the best option. Visually boring, perhaps, but quite readable. If you want to show importance, list them in ranked order.

But enough with the nitpicking. I really, really love the amount of information provided and the presentation and formatting and all of it. It's seriously cool. I'm going to recommend it to my local public library and the library where I work... and any other library who will listen to me. :)

(Edited because English is hard today.)

I really should have posted my comment here, rather than in the Development thread as I originally did. So to remedy that:

I've looked through the sample Syndetics Unbound (will that be LTSU here on Talk?) and like the information, the dynamic interface, and the design. It's an initial impression, but it's a good one.

>8 _Zoe_:

Well, I think the way we do tagmashing might benefit from it. With the ability to tagmash on anything, as was said.

>13 timspalding: I'd have no problem with that, compared to the current tagmash system where we're just typing into an empty box.

But my bigger tagmash priority is to have a nice visual display of the results—I can come up with the search terms on my own.

Just out of curiosity, is it possible to see somewhere which if any of one's reviews have passed vetting and sent out into this (and/or LTFL - I guess it's the same ones?)?

Just out of curiosity, is it possible to see somewhere which if any of one's reviews have passed vetting and sent out into this (and/or LTFL - I guess it's the same ones?)?

It's not exposed on LT.com. Frankly, if we did, people would complain about it. Why this and not that? Etc. (The answer would generally be: Because we haven't gotten to it yet, or because it was automatically excluded for some reason.)

>17 AndreasJ:

Well, I can see why you'd want to minimize the opportunities for people to be disappointed that way. Still, it'd been nice to be able to still my curiosity conveniently.

The tag boxes looked like a change for change sake. But I'll deal with that since I now will be able to figure out (easily, that is) which book to read in a series! Thank you, thank yo.......

wow, none of the links in the blog worked for me, maybe later I'll try again

Sorry, Encore is not available at the present time. Please try again later or contact your librarian
http://hplct-encore.iii.com/iii/encore/record/C__Rb1482234__Shamilton__P0,12__Or...
Error code: 2001

>19 virg144: Not just you, assuming by "links" you're only referring to those that go directly to the examples in the Hartford PL included in the post (links to stuff here on LT are working fine). Looks like something is up with their catalog at the moment.

Yeah, they are trying to upgrade their catalog, and it's not going well. It doesn't have anything to do with Syndetics Unbound. A real pain.

I think the new tag clouds are great, especially in that context. I'd try to filter out tags based on the author's name as well.

(Copied from blog comment)

Woohoo! This looks great! I bet it will help fill coverage gaps in current Syndetics coverage.

Will this be a replacement for the existing Syndetics metadata system or is it an entirely new product?

Will this be a replacement for the existing Syndetics metadata system or is it an entirely new product?

Best answer—yes!

It's half an upgrade to Syndetics, and half a new product. The Syndetics elements are all there and will be familiar, and the LTFL ones added. But everything's been rethought from the ground up, combined in a single, simple solution, and made dynamic and attractive in a way it never was before.

>24 timspalding: Someone's been practicing writing marketing materials.

>25 PhaedraB:

https://www.youtube.com/watch?v=URp-Dbhr7EU

>26 Crypto-Willobie: :-D

>25 PhaedraB:

Tell me about it. But I believe every word of it—that's the difference. :)

>26 Crypto-Willobie:

That's such a classic track, and jarring that I get the blaring YouTube caution. A pretty innocuous piece, given what else is on offer on YouTube but without any other type of warning!

>28 timspalding:

I'd say Waits believes every word he says there, too ... but I get your point. The narrator in the song is a shyster, to be sure.

I can't see it unless I want to 'join', which I have no intention of doing.

First time I've gotten that.

I'm actually absurdly proud of Syndetics Unbound. The original idea was to smoosh the two services together. But as we got into it, we kept moving the bar higher and higher. It drove project management crazy, but we did it—and were only late by a week or two.

The end result is simply the best catalog enrichment out there. I believe that, and I believe it's not even close. It blows the competition out of the water. Can I say that? Fuck it. I'll say it.

Now, will it sell? I don't know. One programmer we once had said that, if we made the perfect product for libraries, nobody would buy it. That is, you have to make the same shit other people are making, but with some tweak that may or may not relate to making it better. But a really NEW product doesn't work.

In this case, we broke one basic rule--that everything has to be bought in little pieces. Previous enhancements were menus. It didn't cost us or ProQuest more to provide each additional feature, but everything was monetized by just HOW many you turned on. So every customer would pick and choose, and somehow that added up to a price. Customers were trained on this monetization model, and came to think it wasn't just a model but a truth.

SU breaks that model. We're giving you everything, for much less than it would have cost to buy it all separately. If you don't want part of it, fine. We expect people won't want every enhancement, and that's fine. But it's not clear that libraries will understand this, when they've been trained to think something should cost 5% less if you are only going to use 19 out of 20 elements of a product. And if you give people something for 1/3 of what it "should" cost, and they don't want part of it, they may expect it to be 5% off from 1/3…

>31 timspalding:

I like that approach. I have little experience or evidence of what strategy works, so my opinion is from a general outlook, not from business. Which is to say, my opinion is idealist.

That said, a possible benefit to your approach is that users will learn what they want from having it, and not assuming a pure rational decision-maker can choose and pay only for what they need. Another benefit: price discrimination is good in theory, but can easily come apart at the seams. Witness what United Airlines is attempting to do with paying for overhead luggage ....

>30 MarthaJeanne: Apparently it's age-restricted due to content, so signing up with a birthdate is the only way for them to "verify" things (which is senseless as anyone can put any birthdate anyway but hey). I have old google accounts I haven't touched in years, not about to touch them now just to watch youtube, lol.

It will be interesting. Syndetics has some limitations in coverage especially with materials published outside of Anglo-America. I know a bunch of places have published on multi-source enhancement scripts (whether they strictly adhere to TOS or not). Have you done any studies on how much broader coverage is with the combined data?