First look: Semantic Web App "Twine"

[podtech content=http://media1.podtech.net/media/2007/12/PID_013179/Podtech_TWINE.flv&postURL=http://www.podtech.net/home/4721/twine-semantic-web-tool-revealed&totalTime=3546000&breadcrumb=f28cfc18758a4308be50f5808e5c85e0]

You’ve heard the hype about Web 3.0. Er, the “Semantic Web.”

Well, here’s the first look at it with Radar Networks’ Twine, which is like a bookmarking/collecting tool on steroids.

If you only have 10 minutes, here’s a short, edited version of the interview/demo.

In the hour-long version (also embedded on this blog post) CEO Nova Spivak gives the first in-depth public demo of the technology.

Twine is darn cool, and something I’m looking forward to using. But it’s good to see this as a real-world demonstration of what the Web could look like.

Comments

  1. Bookmarking/collecting tool? Groan. Didn’t like 200 of these types of sites die in 1999? And shareware sites are littered with this type of ‘tip calculator’ crapware. Now with a Web 2.0 spin, O frabjous day! Callooh! Callay!

  2. Bookmarking/collecting tool? Groan. Didn’t like 200 of these types of sites die in 1999? And shareware sites are littered with this type of ‘tip calculator’ crapware. Now with a Web 2.0 spin, O frabjous day! Callooh! Callay!

  3. Twine is really a tool that helps you keep track of your interests in a smarter way. Bookmarking is only part of what it does. You can also add and author information to share with others. And Twine learns. It’s smart. It self-organizes and gives you recommendations as it learns about your interests.

  4. Twine is really a tool that helps you keep track of your interests in a smarter way. Bookmarking is only part of what it does. You can also add and author information to share with others. And Twine learns. It’s smart. It self-organizes and gives you recommendations as it learns about your interests.

  5. Like Christopher Coulter, I haven’t understood where was the smart in this product. It’s more a free web browser add-on that you want to use to MANUALLY populate a silo database (which is a bad idea since your ontology is guaranteed to be different than someone else’s).

  6. Like Christopher Coulter, I haven’t understood where was the smart in this product. It’s more a free web browser add-on that you want to use to MANUALLY populate a silo database (which is a bad idea since your ontology is guaranteed to be different than someone else’s).

  7. Three comments so far and two of them negative, and that is without even trying the software.

    Calling it bookmarking is a bit of underselling, but I guess it is fair at this point. This is just a foundation. What is built on that should be more interesting. Foundations are only cool if you have seen the blueprints for the whole building.

  8. Three comments so far and two of them negative, and that is without even trying the software.

    Calling it bookmarking is a bit of underselling, but I guess it is fair at this point. This is just a foundation. What is built on that should be more interesting. Foundations are only cool if you have seen the blueprints for the whole building.

  9. Yeah we don’t call it a bookmarking tool at all actually. We call Twine a smart service for tracking and sharing knowledge about interests.

    The key is that Twine is smart. It’s not just another dumb info tool. It understands the meaning of unstructured and structured information you add to it, it organizes automatically, and it learns and begins to recommend stuff. We use natural language understanding, as well as statistical and graph approaches to figure out what stuff is about, and how it is related to other things. So as you use it, you are teaching it, and then it starts getting better. It’s one of the first information tools that actually gets better as you add more information too it, instead of getting worse.

    Twine is built on an entirely new platform, based on the Semantic Web. So all the information you add to is actually connected to an ontology under the hood. And it’s extensible. More on that later. We’re planning to begin rolling out to broader audiences in Q1.

  10. Yeah we don’t call it a bookmarking tool at all actually. We call Twine a smart service for tracking and sharing knowledge about interests.

    The key is that Twine is smart. It’s not just another dumb info tool. It understands the meaning of unstructured and structured information you add to it, it organizes automatically, and it learns and begins to recommend stuff. We use natural language understanding, as well as statistical and graph approaches to figure out what stuff is about, and how it is related to other things. So as you use it, you are teaching it, and then it starts getting better. It’s one of the first information tools that actually gets better as you add more information too it, instead of getting worse.

    Twine is built on an entirely new platform, based on the Semantic Web. So all the information you add to is actually connected to an ontology under the hood. And it’s extensible. More on that later. We’re planning to begin rolling out to broader audiences in Q1.

  11. Hmm, I posted a comment with my wordpress account and it’s not here nor in my comments area. But it shows up on this page when I log in. V. odd.

    Anyway, thanks for the video Robert, keep em coming.

    *I’m new to wordpress.

  12. Hmm, I posted a comment with my wordpress account and it’s not here nor in my comments area. But it shows up on this page when I log in. V. odd.

    Anyway, thanks for the video Robert, keep em coming.

    *I’m new to wordpress.

  13. Thanks for the video Robert, keep em coming.

    By the way I posted a comment with my wordpress account and it’s not here nor in my comments area. But it shows up on this page when I log in to wp as comment 3 from booktagger. Verry odd.

    *I’m new to wordpress.

  14. Thanks for the video Robert, keep em coming.

    By the way I posted a comment with my wordpress account and it’s not here nor in my comments area. But it shows up on this page when I log in to wp as comment 3 from booktagger. Verry odd.

    *I’m new to wordpress.

  15. I watched the full hour video. The concept and the approach that Twine is taking is absolutely the sort of capability that we’ve been needing for our community. I co-chair the Federal Semantic Interoperability Community of Practice. We have following open collaborative approaches using conventional wikis for several years now. The communities we work with range from financial systems, to emergency response, to nanotechnology, to geospatial ontology working groups, to net-centric operations, to semantic service oriented architecture, to community statistics, to electronic health records, and more. Nova’s points about needing better ways to organize, share, and mobilize the information and knowledge that emerges from these techno-socio collaborative activities resonate with us. We’re looking forward to being able to tie it all together. Great presentation!

  16. I watched the full hour video. The concept and the approach that Twine is taking is absolutely the sort of capability that we’ve been needing for our community. I co-chair the Federal Semantic Interoperability Community of Practice. We have following open collaborative approaches using conventional wikis for several years now. The communities we work with range from financial systems, to emergency response, to nanotechnology, to geospatial ontology working groups, to net-centric operations, to semantic service oriented architecture, to community statistics, to electronic health records, and more. Nova’s points about needing better ways to organize, share, and mobilize the information and knowledge that emerges from these techno-socio collaborative activities resonate with us. We’re looking forward to being able to tie it all together. Great presentation!

  17. Good Stuff. I will contact with you guys to
    see if it is possible to promote Twine in China.
    A lot of excellent technologies have been
    adopted in Twine. At first step, I will recommend this
    presentation to my colleagues around here.

  18. Good Stuff. I will contact with you guys to
    see if it is possible to promote Twine in China.
    A lot of excellent technologies have been
    adopted in Twine. At first step, I will recommend this
    presentation to my colleagues around here.

  19. So, each Twine has an email address and users can add data simply by sending email to it. That’s a very convenient way of updating data.

    With email comes the spam. How they are going to prevent it – is the email address a secret, like a hash function of something (“dgsf4545asdfah5@twine.com”)?

  20. So, each Twine has an email address and users can add data simply by sending email to it. That’s a very convenient way of updating data.

    With email comes the spam. How they are going to prevent it – is the email address a secret, like a hash function of something (“dgsf4545asdfah5@twine.com”)?

  21. Not interesting huh?

    I have a different prediction. I believe Twine is one of a very few new services that promises (I have to say promises because they haven’t let me into the beta, despite my pleas) to deliver on weak tie social network value props. If they can avoid over indulging in support for strong-tie relationships (one of the reasons Facebook has so little relevance to knowledge professionals), we might have a winner here. We might be looking at the first information-worker relevant social/semantic service.

    The category, if not the service itself, is game changing.

  22. Not interesting huh?

    I have a different prediction. I believe Twine is one of a very few new services that promises (I have to say promises because they haven’t let me into the beta, despite my pleas) to deliver on weak tie social network value props. If they can avoid over indulging in support for strong-tie relationships (one of the reasons Facebook has so little relevance to knowledge professionals), we might have a winner here. We might be looking at the first information-worker relevant social/semantic service.

    The category, if not the service itself, is game changing.

  23. I’m a contributing author to SemanticReport.com and an extremely active Twine private beta tester, attended my first IJCAI conference before California was a state, ran the web agents session during the First International Conference on Autonomous Agents. Was getting street cred in this space when the only three companies that mattered were LMI, Teknowledge and Inference. All this being said, Twine isn’t just a cute take on Web 3.0. Twine is much more than this. In simple terms, it marries AI with social networking, allows users to painlessly expand their social network(s).

    The semantic processing helps to eliminate the necessity for tagging. Let’s face it, tagging sucks. This doesn’t mean that a user can’t add tags, it simply means that they don’t have to add tags. Twine learns from what a user is reading/viewing, be it from bookmarks, documents (think Scribd), images, videos, all types of content. And not only can Twine recommend related content, but related Twines — and related users. This, to me, is the beauty of Twine.

    Facebook, Plaxo, LinkedIn, pick your favorite social network. Fact is, they’re really nothing more than glorified contact management systems. Not a bad thing, but certainly not the Holy Grail. Twine goes much further, letting me know who SHOULD be in my social network. This is invaluable for a host of reasons. I want new contacts, not a fancy way of handling my existing contacts. I want to meet new people, kindred souls, those with similar interests. I don’t want to have to search to find these people; I want a system to tell me who these people are — and I want solid recommendations from the system.

    Content-based semantic processing is a solution to this problem/need. And this is the beauty of Twine. Make tagging unnecessary. Introduce new people. (Think online dating, too, especially with the recent $500 million acquisition of Various Inc.) Automatically create and expand one’s social networks.

    In comparison to Twine, Facebook seems so retro.

  24. I’m a contributing author to SemanticReport.com and an extremely active Twine private beta tester, attended my first IJCAI conference before California was a state, ran the web agents session during the First International Conference on Autonomous Agents. Was getting street cred in this space when the only three companies that mattered were LMI, Teknowledge and Inference. All this being said, Twine isn’t just a cute take on Web 3.0. Twine is much more than this. In simple terms, it marries AI with social networking, allows users to painlessly expand their social network(s).

    The semantic processing helps to eliminate the necessity for tagging. Let’s face it, tagging sucks. This doesn’t mean that a user can’t add tags, it simply means that they don’t have to add tags. Twine learns from what a user is reading/viewing, be it from bookmarks, documents (think Scribd), images, videos, all types of content. And not only can Twine recommend related content, but related Twines — and related users. This, to me, is the beauty of Twine.

    Facebook, Plaxo, LinkedIn, pick your favorite social network. Fact is, they’re really nothing more than glorified contact management systems. Not a bad thing, but certainly not the Holy Grail. Twine goes much further, letting me know who SHOULD be in my social network. This is invaluable for a host of reasons. I want new contacts, not a fancy way of handling my existing contacts. I want to meet new people, kindred souls, those with similar interests. I don’t want to have to search to find these people; I want a system to tell me who these people are — and I want solid recommendations from the system.

    Content-based semantic processing is a solution to this problem/need. And this is the beauty of Twine. Make tagging unnecessary. Introduce new people. (Think online dating, too, especially with the recent $500 million acquisition of Various Inc.) Automatically create and expand one’s social networks.

    In comparison to Twine, Facebook seems so retro.

  25. If Twine lets itself and is capable, I think it could become the tagging engine for the interweb. People will push their content into it, take on lots of standardized tags, and pull it out the other side, “enriched” and ready for its final use, whether that be publication or whatever.

    For instance, what if a blog created a Twine for itself? Its author could stop having to worry about being consistent in tagging his posts; instead, he would just let Twine do the job. As time passed, moreover, Spivack suggests that Twine may get better at the job.

    Or what if a what newspaper, maybe a big one like the New York Times, created a Twine for all its content, maybe even all its historical content? Could Twine replace the set of librarians who currently create the metadata Dave Winer’s been doing fun stuff with?

  26. If Twine lets itself and is capable, I think it could become the tagging engine for the interweb. People will push their content into it, take on lots of standardized tags, and pull it out the other side, “enriched” and ready for its final use, whether that be publication or whatever.

    For instance, what if a blog created a Twine for itself? Its author could stop having to worry about being consistent in tagging his posts; instead, he would just let Twine do the job. As time passed, moreover, Spivack suggests that Twine may get better at the job.

    Or what if a what newspaper, maybe a big one like the New York Times, created a Twine for all its content, maybe even all its historical content? Could Twine replace the set of librarians who currently create the metadata Dave Winer’s been doing fun stuff with?

  27. @Josh Young, of course, human editors can often do a better job than “machine” editing. But I’ve learned from using Copernic Summarizer that the “Concepts” (think “tags”) assigned to a piece of content are often better than author-provided tags; I’ve noticed this with experiments I’ve done with journal articles with author-provided “Keywords”. So with the New York Times, it may not be an either/or, but and/both. But that’s the New York Times. Few media outlets have their resources or their clout to garner third-party support.

    Enter Twine. As you have suggested, what about a “Twined” blog? This is in the works. How about a universal criteria for the selection of linguistic units? An individual can add their own tags, if they prefer, but Twine can assign tags from a universal corpus. (It already does this in the private beta.) The result: IDA (intellectual data analysis) rather than run-of-the-mill text or data mining.

    Twine is game changing. It’s truly much more than just Web 3.0; it’s Interactions 3.0. And compared to the other Semantic Web players, Twine is a clear leader with social networking features; it’s a priori to what Twine is all about, to their corporate mindset. Their goal doesn’t seem to be to destroy Ask and then take on Live, Yahoo and ultimately Google. For this, there is Powerset, Hakia.

  28. @Josh Young, of course, human editors can often do a better job than “machine” editing. But I’ve learned from using Copernic Summarizer that the “Concepts” (think “tags”) assigned to a piece of content are often better than author-provided tags; I’ve noticed this with experiments I’ve done with journal articles with author-provided “Keywords”. So with the New York Times, it may not be an either/or, but and/both. But that’s the New York Times. Few media outlets have their resources or their clout to garner third-party support.

    Enter Twine. As you have suggested, what about a “Twined” blog? This is in the works. How about a universal criteria for the selection of linguistic units? An individual can add their own tags, if they prefer, but Twine can assign tags from a universal corpus. (It already does this in the private beta.) The result: IDA (intellectual data analysis) rather than run-of-the-mill text or data mining.

    Twine is game changing. It’s truly much more than just Web 3.0; it’s Interactions 3.0. And compared to the other Semantic Web players, Twine is a clear leader with social networking features; it’s a priori to what Twine is all about, to their corporate mindset. Their goal doesn’t seem to be to destroy Ask and then take on Live, Yahoo and ultimately Google. For this, there is Powerset, Hakia.

  29. Super Cool. Twine Rocks. Bravo Nova.. I clearly see this as much more than a book marking tool etc. I dont know how and why – assuming people have seen the video – are underestimating the power of Twine !

    My take on this Twine stuff goes like:

    (A) Emails : Piles if you dont File.
    (B)Content misses Intent – Twine -> ties Content & Intent.

    (C) Tell I ->Click It -> Forget it –> Twine goes beyond Integration. Its automation

    Thats my 2 cents as a former Marketer :)

    Cheers
    ajay

    CEO
    nomolisa.com

    PS: Nova: give me beta access..

  30. Super Cool. Twine Rocks. Bravo Nova.. I clearly see this as much more than a book marking tool etc. I dont know how and why – assuming people have seen the video – are underestimating the power of Twine !

    My take on this Twine stuff goes like:

    (A) Emails : Piles if you dont File.
    (B)Content misses Intent – Twine -> ties Content & Intent.

    (C) Tell I ->Click It -> Forget it –> Twine goes beyond Integration. Its automation

    Thats my 2 cents as a former Marketer :)

    Cheers
    ajay

    CEO
    nomolisa.com

    PS: Nova: give me beta access..

  31. Just what we needed: More “semantic” hype.

    These guys are not anywheres in the vicinity of semantic validation with their glorified bookmarks.

    It’s amazing that this PR gets pumped out month after month, year after year. The actual “progress” being made consist of still greater numbers of useless toys and re-compilations of Wikipedia.

  32. Just what we needed: More “semantic” hype.

    These guys are not anywheres in the vicinity of semantic validation with their glorified bookmarks.

    It’s amazing that this PR gets pumped out month after month, year after year. The actual “progress” being made consist of still greater numbers of useless toys and re-compilations of Wikipedia.

  33. [...] Scobleizer – First look: Semantic Web App “Twine” Video interview on Radar Networks’, Twine. A semantic information manager. (It’s more interesting than it sounds!) (tags: twine collectiveintelligence semantic web tagging information knowledgemanagement data database api standards socialnetworking collaboration wikis videos “semanticgraph”) [...]

  34. I got into the Twine beta last week and I’ve been playing around with it since then. It’s pretty neat, but so far I don’t see any functionality besides simple bookmarking and aggregating documents. Everything I hear and read about it says it should do a lot more than this. From how Nova describes it it sounds awesome.

    Maybe I’ll just keep adding data to it and see if it eventually starts organizing it or doing something useful other than scanning pages and documents for keywords.

  35. I got into the Twine beta last week and I’ve been playing around with it since then. It’s pretty neat, but so far I don’t see any functionality besides simple bookmarking and aggregating documents. Everything I hear and read about it says it should do a lot more than this. From how Nova describes it it sounds awesome.

    Maybe I’ll just keep adding data to it and see if it eventually starts organizing it or doing something useful other than scanning pages and documents for keywords.

  36. I just signed up so I cannot comment yet on Twine. But just one thought though, are the info & links we leave, going to or be used by a third party?
    Cheers BC

  37. I just signed up so I cannot comment yet on Twine. But just one thought though, are the info & links we leave, going to or be used by a third party?
    Cheers BC

  38. Hi

    Our approach at Imindi (An application built by a team of Phd computer and neuro scientists) is to take a “mind” approach to helping people to construct their own “mind maps” of connected thoughts and information on any subject. These Mind Maps are little semantic webs that work “As we think” and the “Thought Engine” (Semantic Graph) at the core of Imindi enable Like Minded people (Social Graph) to connect and combine the Thoughts, Information and even Create Knowledge. At the core, all the semantic linkages from every body`s public mind maps collape on themselves to form essentially one global mind map.

    However, the closest vision to what we have built was not that of the Semantic Web, but that of the Memex by Vanevver Bush – but as we build a layer of thought and meaning over the information on the web…I would suggest that in addition to Twine we are also very much a Semantic App worth watching.

    http://www.imindi.com/journeys/382-semantic-web/maps/3195155-semantic-web

  39. Hi

    Our approach at Imindi (An application built by a team of Phd computer and neuro scientists) is to take a “mind” approach to helping people to construct their own “mind maps” of connected thoughts and information on any subject. These Mind Maps are little semantic webs that work “As we think” and the “Thought Engine” (Semantic Graph) at the core of Imindi enable Like Minded people (Social Graph) to connect and combine the Thoughts, Information and even Create Knowledge. At the core, all the semantic linkages from every body`s public mind maps collape on themselves to form essentially one global mind map.

    However, the closest vision to what we have built was not that of the Semantic Web, but that of the Memex by Vanevver Bush – but as we build a layer of thought and meaning over the information on the web…I would suggest that in addition to Twine we are also very much a Semantic App worth watching.

    http://www.imindi.com/journeys/382-semantic-web/maps/3195155-semantic-web