Comments for Twine: The First Mainstream Semantic Web App?

Twine: The First Mainstream Semantic Web App?

2007-10-19T04:00:00Z

On Friday Radar Networks is announcing a new Semantic Web application called Twine. Founder Nova Spivack showed me a demo today of the new app, which he described as a "knowledge networking" application. It has aspects of social networking, wikis, blogging, knowledge management systems - but its defining feature is that it's built with Semantic Web technologies. Spivack told me that Twine aims to bring a usable and scalable interface to the long-promised dream of the Semantic Web.

Spivack went as far as to claim that Twine will be "the first mainstream Semantic Web application" - and it's certainly fair to say that we've heard lots of theory about the Semantic Web ever since Tim Berners-Lee defined it, but as yet there have been very few large scale success stories (if any). Will Twine finally be the Semantic Web app that breaks through? Let's find out more...

First some background: Nova Spivack has an illustrious history in the Semantic Web and AI business, having worked for both AI legend Ray Kurzweil and tech guru Danny Hillis (Thinking Machines). The genesis for Twine, said Spivack, came from an R&D project about 5 years ago, which turned into a research project, then a Series A round with Microsoft co-founder Paul Allen in 2006. As of now the Twine team is 30 people working from San Francisco -- and they're finally ready to unveil their new mainstream Semantic Web product.

What is Twine?

The aim of Twine is to enable people to share knowledge and information. At first glance it is very much like Wikipedia, but there is a whole lot more smarts to the system. Spivack described it to me as "knowledge networking"- i.e. it aims to connect people with each other "for a purpose". It's not based around socializing, but to share and organize information you're interested in. Using Twine, you can add content via wiki functionality (there are many post types), you can email content into the system, and "collect" something (as an object, e.g. a book object). The screenshots below show of this in action -- note that the product itself isn't available just yet, as it's in private testing.

]]> Other features of Twine include: RSS feeds to track all kinds of things (topics, events, search, etc); commenting and viewing related things, sharing tags, and more. Also, and Marc Canter will like this, Twine users will be able to import and export their own data. Nova said that Twine will be an open platform - there will be a SPARQL API and a REST API.

Semantic Graph

Where Twine is differentiated from the likes of wikipedia is that its underlying data structure is entirely Semantic Web. Spivack told me that the following Semantic Web technologies are being used: RDF, OWL, SPARQL, XSL. Also he said that they plan to use GRDDL in the near future. Spivack had an interesting term for what Twine is doing with Semantic Web technologies, riffing off the Facebook Social Graph. Spivack is calling Twine a "Semantic Graph", which he says will map relationships to both people and topics. So Twine's Semantic Graph actually integrates the Social Graph. Spivack said that his company has patents pending on this.

Who will use it?

So who is Twine aimed at? Spivack said that it's aimed at professionals and teams. Also he said content providers are expressing interest, because their data can be turned into Semantic Web data and re-used.

As for the business model, it will be advertising and also subscriptions (for higher capabilities). The advertising part won't be in the first release and Twine hasn't yet decided how to run that - e.g. they may use a single ad network provider, or (as Facebook is considering) create their own ad network.

Conclusion

Overall, while the app isn't ready yet for the public, I was impressed with what I saw in Nova's demo. The proof will be in the pudding regarding whether this will be the first mainstream Semantic Web app - i.e. how much uptake it gets and whether there will be good use cases for all this semantic data. But it certainly looked like a usable and slick system - and one I'm looking forward to playing with.

Screenshots

Click each picture to see full-length screenshots.

Comment from Yihong Ding on 2007-10-18

2007-10-19T04:20:43Z

Very interesting. I cannot wait for the formal release and get to try it.

Comment from alien on 2007-10-18

2007-10-19T05:48:10Z

i'm building something similar, and I'm just one guy. I have called these kinds of things webswarming, or memetic ecosystems.. I pray Twine has an API (though I consume feeds, and surely they must feed..) www.seekng.com, www.lorestorm.com

Comment from peter royal on 2007-10-18

2007-10-19T06:07:06Z

@alien, yes twine has an API.

you'll be able to get atom feeds, and there is a data-oriented API as well.

Comment from javaeater on 2007-10-18

2007-10-19T06:32:12Z

Never seen such complicated system ever won. User is lazy; some of my friends now even don't want to tag in del.icio.us any more.

Keep it in mind team or just type www.google.com and think on it.

Comment from Chris Messina on 2007-10-18

2007-10-19T06:45:53Z

Hmm, no mention of microformats. If there's support for GRDDL, I'd hope we'd see some microformats in there as well?

Comment from Derek on 2007-10-18

2007-10-19T06:49:22Z

Sounds great, possibly similiar to Lijit in many ways (apart from the semanticism)?

I was disappointed to see some marketing Gobbledygook in the Twine blurb that turned me off somewhat..."Twine, a new service that gives users a smarter way to share, organize, and find information with people they trust."

That says nothing, or at least, nothing that 350 other startups are saying. Please people, read about the new rules of PR, if you want to attract early-adopters don't talk to them like an infomercial audience.

Frankly, I got far more sense from the R/WW article. The fact that R/WW carries some weight makes me consider exploring Twine despite the poor promotion. C'mon Twine.

Comment from Yakov on 2007-10-19

2007-10-19T07:47:52Z

I am dubious if it would go mainstream anytime

Comment from Aaron Mentele on 2007-10-19

2007-10-19T07:52:26Z

How's freebase doing? Weren't they the first semantic web app?

For the record, I'm pulling for Twine - sounds exceptionally cool (been following Nova's blog for a while.)

Comment from Lars Teigen on 2007-10-19

2007-10-19T08:35:09Z

This looks very promising. Signed up for the beta. Can't help but to mention that Second Brain is working on something similar, though without the semantic connections.

Comment from bernard lunn on 2007-10-19

2007-10-19T12:12:45Z

This looks like a great fit for what a big need I am seeing. I call them Research Networks, but same difference. Don't think consumer apps - think scientific, technical, medical. I am also enthused about Freebase but have seen nothing since the March announcement.

Comment from Ivan Plestina on 2007-10-19

2007-10-19T13:04:08Z

@ #5
Why would you need microformats with RDF that is far more advanced and open API?

On topic:
It really looks promising. What I'm interested in is if it will support closed groups and, more importantly, closed data for those groups? It would create a nice platform for companies to keep their data on.

Comment from Shelley on 2007-10-19

2007-10-19T13:50:28Z

@11,

Ivan, Chris Messina's comment was valid. Firefox 3 has built in capability to 'read' certain microformats, such as those for events. Publishing the data in microformat form directly in the page is not counter to the use of RDF. In fact, with this application, it would be complementary to the use of the technologies listed.

Comment from Ivan Plestina on 2007-10-19

2007-10-19T14:17:33Z

Indeed, but already now Operator extension for FF can read embedded RDF. How good is native support for it in FF 3 I do not know but I see no reason why it would lack support for it when it comes out.

I just see no point implementing microformats looking from Twine's perspective. Why would they mess with two technologies if they can have same results with just one?

Comment from peter royal on 2007-10-19

2007-10-19T14:38:29Z

@ #11 -- yes, private groups are baked in. there is a strong ACL layer on top of everything.

@ #8 -- yeah, free base was the first Semantic Web app if you take the word 'semantic' literally (oh, semantics!). They aren't built on top of the notion of semantics as put forth by the W3C though, so that's a key differentiator (well, that and what the applications are!)

@ #5 -- as grddl is "just" xslt at the moment, quite trivial to grab microformats out via the same mechanism :)

Comment from Shelley on 2007-10-19

2007-10-19T16:16:00Z

@14

"yeah, free base was the first Semantic Web app if you take the word 'semantic' literally (oh, semantics!). They aren't built on top of the notion of semantics as put forth by the W3C though, so that's a key differentiator (well, that and what the applications are!)"

What an extremely curious thing to say. What exactly is the notion of semantics as put forth by the W3C, and how does Twine differ? I didn't quite catch that from the writings on this release.

Comment from Shelley on 2007-10-19

2007-10-19T16:17:32Z

Actually, I may have answered my own question:

Is it because the W3C promotes the human application of metadata, and Twine is based on the use of natural language processing to derive metadata?

Comment from Shelley on 2007-10-19

2007-10-19T16:22:49Z

Sorry, if I could edit comments, I would merge these three..

@13, Ivan: "I just see no point implementing microformats looking from Twine's perspective. Why would they mess with two technologies if they can have same results with just one?"

Because the benefit accrued to the act could outweigh the rather trivial amount of effort expended in order to provide the same data in two different formats. There are applications that can process microformats, ones that can process RDF, and ones that can process both. Why limit one's sparkly new app?

I'm a strong fan of RDF, and frequently poke fun at the microformats folks, but in this case, providing the microformat formatted data would be so trivial it becomes a case of, "Eh? Why not?"

Comment from Nick Dynice on 2007-10-19

2007-10-19T16:58:22Z

It seems interesting but I have a problem an app that has both data portability but then also a patent. So once I have the data out I can't put it into another application that does the same thing because such an app is hampered by patents.

Comment from mike hastings on 2007-10-19

2007-10-19T17:44:53Z

I am wondering if there are any enterprisey business apps that take advantage of semantic web technologies already.( I mean RDF\OWL as data format sounds useful and appropriate but most of biz apps are built on RDBMS and layering RDF Store on top of RDBMS apps may lead to more issues.) Any pointers on how to approach it?

thanks
mike

Comment from Eric Monse on 2007-10-19

2007-10-19T18:59:49Z

It looks pretty sweet. But does the world need another one of these? - Eric Monse

Comment from Antoine of MMM/Brighthand on 2007-10-19

2007-10-19T19:43:36Z

I like the idea behind this, but (just from a personal standpoint), something like this would come in a good deal more handy if it plugged into the already existing social networks (like a layer above it) and into my mobile (appending the address entires with the connection-context content).

Specifically towards the mobile end of things, this could be really good; but I don't see it hitting anything near what the majority of users would want it for, as its yet to solve that "dang it I need that" aspect of the way of lives for not-so-web20ish-users.

Nevertheless, its a great idea, and glad they and others are pushing this out.

Comment from Mikey on 2007-10-19

2007-10-19T20:15:38Z

Nice review I have also just reviewed Twine.com on www.WhichWebsite.com check this and other website related news at http://www.WhichWebsite.com

Comment from Justin on 2007-10-19

2007-10-19T20:59:49Z

Does anyone know how to get an invite to Twine. I would really love to check this thing out for myself.

Thanks

Comment from peter royal on 2007-10-19

2007-10-19T21:05:50Z

@ 16 - by "w3c semantics" i primarily mean building on top of the RDF data model.

@ 23 - sign up for the beta list at http://www.twine.com

Comment from Dale on 2007-10-19

2007-10-20T03:37:48Z

Patent Pending? These are all open standards technologies. What's to patent?

Comment from Sherwin Shao on 2007-10-20

2007-10-20T23:43:22Z

Is this just more vapourware from the great promise of Web 3.0? Why will this grand vision of Web 3.0 involving intelligent machines never happen? Because humans will never be surpassed by machines in judgemental intelligence. To be interested in certain things, that is a uniquely human trait. To belong to groups, another human trait. Because a machine-generated solution, regardless of how advanced, cannot replace humans. In the end, only humans can help other humans.

So What‚Äôs Next?

Different words have different meaning to different people. Different people are identified by different demographics. There‚Äôs no need for the machine to understand all that meaning stuff. As long as people understand, and communicate with the smallest possible unit of related meaning. Which is the question and answer.

By simply matching questions with other similar questions, and group the answers from all the experts, users are given access to the best possible knowledge available from all perspectives. So you can learn at a much faster rate, bypassing the uninteresting noise for the interesting nuggets of knowledge.

The next generation of the web will make use of what we like, what we know, and what we‚Äôve done, to give us what we need. Based on our recent search history. Based on our demographics. Based on our ratings, questions and answers.

Given your usage, the system should know what you‚Äôre interested in, and can show you questions based on the community you naturally belong to. So you will not have to avoid people you find annoying‚Ä¶ the system will segregate you naturally. Also, other people similar to you will, through their ratings, constantly be finding things that are interesting to you. This system is adaptive, so that as your interests change, your search results change with you.

Try http://www.helpglobe.com. You'll see what I mean.

Comment from Nova Spivack on 2007-10-20

2007-10-21T03:49:22Z

Hey, just in case there is any doubt, our philosophy is to use any and all data and metadata we can see -- including microformats, etc. We're using the semantic web but we're not religious fanatics about it -- we're pragmatists mainly. However, we're also a small team so we can't do everything at once. It will take time to roll out support for all the things on our list. But we certainly are aware of, and open to, the ideas in the suggestions above.

Comment from Dale on 2007-10-21

2007-10-21T07:51:41Z

Nova,

Can you address my question on patents? What are you attempting to patent since you are using existing semantic web technologies? Doesn't it defeat the purpose of community by creating a proprietary solution?

Comment from Major Tom on 2007-10-22

2007-10-22T14:14:52Z

I look at this from two points of view. Knowledge to share versus knowledge to use on one's own project.
I think it would be good to have a fully searchable, linked, knowledgebase. I just wish most Open source projects would see what is out there and contact other projects that are "close" and see if collaboration could be attained.
Wikipedia combined with Twine since there is a lot of data in their DBs. Project Gutenburg, and then there is another that is trying to use Captcha with old documents to help to digitize those documents.
There is a lot of stuff out there. Can no one attempt to combine this? Even Google's technology cannot solve this, we need visionaries and teks to pull this all together.
Give me a couple $10M and I'll do it. Its not hard, just time consuming.
You are correct to monetize this in order to sustain it.
Good luck.

Comment from Matt Ellsworth on 2007-10-22

2007-10-22T14:53:13Z

looks great - i'll be checking that out.

Comment from Qunicy Jones on 2007-10-22

2007-10-23T02:38:58Z

The article fails miserably in stating

a) how the underlying semantic web-based backbone provides any benefits to users

b) how new information is integrated into the semantic web-based backbone

Useless

Comment from Chris Rijnders on 2007-10-23

2007-10-23T09:01:30Z

It is very unclear in the article how Twine will keep seperated:
1) "raw" data
2) information
3) the knowledge extracted from the information
4) the meta overview generated from the knowledge

If they expect users to organise and update these differences themselves I think it probably will quickly become quite an unmanagable mess and (from the point of view of engineering) not a tool that adds value. What users will want to do is just "dump" the data and then have the software sort out the higher-level connections.
Such things are not possible yet? Well then the Semantic web is not here yet.

Comment from Jim Kern on 2007-10-24

2007-10-24T14:02:01Z

An API and ability of the programmers to use the feeds, is key. Could benefit all the smaller interest groups that are a little too far off the radar to be addressed direcly.

Comment from ryanol on 2007-10-24

2007-10-24T14:09:59Z

funded by Paul Allen.....anyone else thinking -
Outlook 3-dot-oh

Comment from pawel lubczonok on 2007-10-28

2007-10-29T05:54:08Z

At the moment the semantic web technologies such as Twine are expressing surface semantics, i.e. no deep representation is present. This will mean that it will be only a staging point onto deep semantic representation systems.

Can one run enterprises using this semantic web, by representing full meaning of knowledge of enterprise.

For last five years we have run enterprises entirely on deep semantic/knowledge representation machine in off line setting. Early next year we our offering will be made available on the web under thoughtexpress.com.

Anyway, our experience with deep semantic (knowledge) representation shows us that amongst many benefits, the IT/KT world becomes accessible to non IT person. For example, it is not possible to have DIY enterprise. In our setting actuaries without any IT training are able to control/develop insurance enterprise at a 5% of time that traditional development structures would take. So, deep knowledge representation is next important step in democratisation of the web.

Pawel Lubczonok

Comment from Mohit Gupta on 2007-10-31

2007-10-31T09:23:49Z

There is a similar application with a stronger focus on Knowledge Vizualization; These guys got a lot of attention at the Imagine Cup Software Design and Innovation Conference at Bangalore organised by Microsoft recently.

More at http://www.projectrecog.com