PaperPort 11 with PaperPort Watson now available

When I wrote my last post (all too long ago now) we had just delivered code that was going to be pressed onto CDs.  Now that PaperPort 11 is available, we’re seeing tons of new users of PaperPort Watson, the version of Watson that comes with PaperPort.  I can’t wait for it to be available in the retail stores so I can take my mom to see the box.

I can’t say enough about the professionalism of the team at Nuance – they were a pleasure to work with.  The exciting new features we were able to deliver with them make Watson even more powerful and give PaperPort users a radically new way to discover information and incorporate it into their document libraries.

With PaperPort and Watson together, you can drag a link from Watson into PaperPort and it will automatically be converted to PDF.  This means that even if the web site changes, you’ll have a copy on your hard drive for later.  Watson also integrates with PaperPort’s desktop search, so documents you’ve stored will surface in Watson when they’re relevant to what you’re doing.

It’s exciting to be able to reach this user community so they can get the benefits of Watson.

What’s coming next?  We’ve got a lot on the table for the next release that comes directly from user feedback.  Please, keep the feedback coming!  I’m just as excited about the features in Watson 2.4 that we’d been planning for what seems like forever—they’re going to really flex the platform and give users the ability to even more easily leverage the information that matters to them.  If you want to try a beta of Watson 2.4, email me.

First Watson RTM; Duet at SAPPHIRE

We just finished testing our first RTM release of Watson, which we’ll be announcing at AIIM on Tuesday.  If you’re attending, shoot me an email, I will be there doing demos.  This our first release to be pressed on CDs (at least that we know about) and the development team rose to the challenge.  We’re really excited about the new functionality and the overall quality of this version of Watson.  We were able to achieve some great product integration in very short order.  Check back later for details.

I’ll also be attending SAPPHIRE in Orlando on Thursday, where I hope to spend some time learning about Mendocino, now officially known as Duet.  This is absolutely where software should be going – there is no reason users should have to jump through so many hoops to do simple things like request vacation time.  With Duet, you can get it done without ever having to leave Outlook.  Of course, we’ve been pushing the same direction with search for some time.  It’s exciting for us to see such big players embrace a similar notion.

Project Cubs

We’ve begun to reach out to power-bloggers in order to hear how we can make Watson more useful to you.  I use Watson while blogging all the time, and find it to be a great way to quickly find supporting links.  Like this discussion on BloggerCon.  If you’d like to participate in designing the next-generation of blogging tools send an email to projectcubs@intellext.com.

Watson 2.3 is here!

Some of you may have noticed we shipped Watson 2.3 last week.  This is a great new version with many enhancements.  One of them is our new drag & drop support.  You can now drag search results from Watson and drop them into the documents you’re working on.  Watson will create a citation like this one:

Draganddrop_3

(Watson found that link for me based on this post I’m writing now.)  You can also drag links from Watson into folders on your desktop, which is a great way to keep a collection of links related to a project you want to visit later.   Just like search doesn’t start when you decide to enter keywords into a search engine,  it doesn’t end when you get that famously long list of search results.  We want to make it easy to incorporate useful information into the work you’re doing, and will continue to deliver features that make this even easier.

Another great feature in Watson 2.3 is our updated Outlook contacts support that includes integration with ZoomInfo.  When you select a contact in Outlook, Watson will do a precise search for the ZoomInfo Web Summary for that contact person.  It will also gather precise results from all of the other sources it is connected with.  This is great for when you’re about to make a phone call and want to quickly get up to speed on one of your contacts.  It’s like having a briefing book delivered to your desktop right when you need it: you get a ZoomInfo web summary like this one, any recent news articles, emails you’ve exchanged with that person, and a link to their home page, if they have one.

This feature shows how Watson can provide even more precise information when knows more specifics about the context in which you are working.  By leveraging structured data in contact records, Watson is able to formulate extremely precise queries, to both structured sources (like ZoomInfo) and unstructured ones (like the Web).  The queries Watson forms based on contacts are very different from those it forms when you are browsing the web.  Because Watson knows that contacts have names, affiliations, positions, and email addresses, it can make sure it finds you information about Martin Lee at IBM instead IBM message board entries posted by Lee Martin.  This feature is a nice demonstration of the leverage you get from paying attention to the semantics, or meaning, embedded in everyday applications, and how Watson can leverage those semantics to deliver information with razor-sharp precision.

We want to hear what you think!  Post a comment or send email to feedback@intellext.com.

-j

Way to go, Andy!

Andy Crossen successfully defended his dissertation, “Building Intelligent Information Systems: Experiences with the ISA Platform,” yesterday.  Andy’s been doing what people are calling “mash-ups” since 1999, way before the term even existed.  His infrastructure, the ISA Platfrom, which we use in Watson, allows developers to easily incorporate multiple web services into a single user experience.  The cool thing is that it doesn’t require the services themselves support Web standards like SOAP and XML.  Developers can write little “wrappers” in XML that provide a programmable interface to any existing web service, even legacy applications and web sites.  The scripting language he built makes this easy – you don’t have to know how to code in C++ in order to add a new service to Watson.  His framework is even capable of automatically reasoning about how to take input and to arrive at the desired output by using the right services together.

During his time as student, he made a bunch of demo apps on top of this.  One of them makes a web site about any book that you scan in with a barcode scanner.  Based on the type of book it will search for different things.  If you scan a fiction book, it will find out about the author.  For a textbook it will see if there are materials online from the publisher.  Smart stuff.

We’re all really excited that Andy has decided to join the Intellext team so he can continue to help us improve Watson!

The search for The Answer

I had the privilege of hearing Google CEO Eric Schmidt speak at the Economic Club of Chicago last week. Matt McCall has a great recap of some of his main talking points.

Eric explained that the Internet has given rise to paradoxes that are new and difficult to understand.  For example, the Internet encourages openness and self-expression, like we see in blogs, while it simultaneously fosters “tribalism,” like hate groups that find each other online and isolate themselves from others.

He went on to talk about how in the future you will be able to buy a Google device that contains a search engine and all the content, so that you can get the answer to your question on a small computer that you carry with you.  Eric thought this would be particularly good for students.  Maybe it would.  In the future, everyone will access everything through Google.

In the same speech, Eric contrasted the old way in which culture was disseminated – through broadcast media, controlled from the top down, with the new way – in which user-generated content spreads like wildfire without any central control.

But Google’s algorithms, the procedures that determine what you see when you type into the search box, are that central point of control.  What you see first is determined by its popularity or “authority” and the algorithms drive towards giving you the “correct” answer.

This is the innovation the Google founders are most well known for.  PageRank, one of the factors Google uses to determine the order of search results, is the idea that links should influence relevance.  Prior to PageRank, most web search algorithms (except Kleinberg’s hubs and authorities algorithm) only considered a page’s contents, not how many other pages linked to it, or what the link said (more evidence for what the page is about).

The effect of PageRank is that the most popular answer is the one that will appear first.  This is good because it’s most likely to be the answer you wanted to see, because, the thought is, all the people on the web have voted by linking and determined that that’s what your search terms mean.

Good until you understand that people rarely browse beyond the first page of search results, not to mention that the ranking can be and is “optimized” artificially.

So, what about all the other answers?  What about all the content that doesn’t make it into the first page of Google results?  What about all of the content publishers who don’t pay companies to optimize their ranking?  What about all of the other factors that influence whether something is relevant to you?

PageRank is the new “old way” – just like the networks control what you see on TV, PageRank controls what you see on the internet in its search for “the answer.”

Perhaps this is the most troubling paradox.

Instead of searching for “the answer,” we should be empowering users to explore the landscape of answers.  Instead of driving to a single result, we should be showing the user the range of contrasting opinions, and giving them tools to discover the ones that make the most sense to them.  The future is not “the answer.”

What is Web 2.0?

Web 2.0 is a catch phrase for the next wave of software that’s all about revolutionary new user experiences that allow users to break the boundaries of traditional, siloed systems and services by integrating them into their lives.  Web 2.0 is about seamless service composition and just-in-time assembly of user experiences.  It’s about putting all the infrastructure built in Web 1.0 to work for the user, on their terms, so they can really take advantage of it.  It’s about integrating the online world into “everyday life.”

A few months ago, I was talking with my friend about how blogs are a lot like notesfiles.  I was a big user of notes, which we had ported from PLATO to SunOS 4.1.3, and then later to Solaris.  Notes was an very basic form of usenet (or netnews).  It was limited because the synchronization and syndication features were limited (all we had was the notes to email gateway).  With usenet, you could post on your local server, and your post would be replicated around the world for people to read.  With notes, your post was stored locally, or on a file server.  To read it you had to have access to the notes file, by accessing a timeshare system via a terminal or terminal emulator.

Notesfiles and blogs are similar.  People post, and you can read their posts, and comment on them.  They have similar social affordances.  They’re an open space to share ideas.  They had cultures, rules, social groups, cliques.  The same is true of blogs and the net in general.  Sure, you can put pictures on blogs, and change the font.  But it’s all about what you post.  Syndication shows that.  Notesfiles even had sequencing tools that allowed users to read all of the new posts at once, much like RSS readers do oday.

Basically everyone I knew seqed and posted to notesfiles.  It was integrated into our work – when we needed to coordinate our activities around a big systems upgrade, or make a quick decision, we checked the notesfile first.   When we wanted to know what people were up to, read funny jokes, talk politics, etc., there was a notesfile for that.  Some of us also knew each other in real life.  But after about 10PM, we interacted on notesfiles.  We were in high school and they made us go to bed.

So why is blogging not like usenet or like notes?  Scale, and ease of use.

There are > 30 million blogs.  We only had 200 notesfiles.  There were about 1000 notesfile readers on our system, and today there are at least as many blog readers as there are blogs.  This of course makes notesfiles safe, familiar, and more controlled.  Like a suburb, or village in the emerging digital landscape.

But why was there such a big jump in use?  Because you don’t have to know how to use a complicated text editor like vi, or even a terminal emulator.  You can blog from your web browser.  Don’t worry, you don’t even have to remember your password.  And now there are tools that make it easy to blog directly from your PC, even when you’re offline.  So don’t worry about a net connection either.

You see, scale and ease of use become a virtuous cycle that lead to huge adoption and growth.  The scale drives innovation in ease of use because value can be placed on increased use. Users integrate these tools into their lives. This is what’s happening with Web 2.0.

Users demand that their services be integrated, not siloed and separated.  The blogging tool should be part of your PC, not some separate application, certainly not through a terminal.  Why is it that email and blogging aren’t integrated?  (You used to be able to use gnus to read both usenet and mail from the same folder hierarchy.)  Services should come together to provide seamless user experiences that leverage information from everywhere.   Users should be free to focus on their goals instead of instrumental tasks, like "posting" or "searching."

We are at the beginning of the end of the idea that we have applications that support instrumental tasks, and moving to a world in which our computing environments adapt to support our goals and activities -- integrating with "everyday life."

To me, that's Web 2.0, and that's what we're trying to do with Watson.  Except instead of doing it on the Web, we're doing it on your desktop, and it's based on what you're doing.  To me, most of the Web 2.0 services miss the boat -- they continue to view themselves as a "destination" instead of embracing the idea that they exist to deliver service to the user, on the user's terms.  Our job is not to create another activity silo.  Instead it's to create experiences that enrich and integrate into everyday life, into what you are doing, so that we can make you better and better at what you do.

-j

SES NYC Interviews

A couple of weeks ago, at SES, Chris Pirillo interviewed me for his show.  Chris and Jake did a bunch of great podcasts from SES about what’s really going on in the search industry, including a fascinating yet infuriating one with "Joe Spammer."  If you don't have time to listen, Lee Odden's got the recap here.

It’s amazing how quickly information spreads through networks.  This list of podcasts is really cool stuff, and it will spread.  It might seem random, but my friend Cameron, did a few studies on it.  His papers and class on information epidemiology are available through his Media Lab site – they make for an impressive overview of “viral” or network effects.

Watson and Microsoft SharePoint

We launched the latest enhancement to Watson today – out-of-the-box integration with Microsoft SharePoint.  Now any SharePoint user can use Watson to tap into their portal content and have it automatically delivered to their desktop.  This is a boon for companies that want to jump-start collaboration and content re-use.  People put systems like SharePoint in place so that people can reuse documents.  Watson helps find opportunities to put those documents to work by letting people know when there’s something relevant out there.  Now users of SharePoint get this automatically.

We chose SharePoint for out-of-the-box integration because so many of our users requested it, and SharePoint provides a robust and scalable enterprise search solution for people who don’t already have something in place. The query language under the hood allows the IT department to customize what’s delivered in Watson by leveraging metadata and taxonomies.  This means that queries dispatched from Watson can deliver more precise results to users.

As always, I’d love to hear what you think about it.  If you’re a Watson Professional user, you can get the update here.  You’ll see the SharePoint integration when you select Tools > Configure Info Sources…  Sorry, this is available for users of Watson Professional only.  If you have the ad-supported version and want to upgrade, click here.

Watson can still be configured to search any searchable source.  Where else do you want to see Watson deliver content from...any suggestions?

-j

Every time we go to New York we buy more equipment

Last week I was out attending Search Engine Strategies in New York. I got the chance to meet Richard Edelman, who had some fantastic ideas for us.  The show frankly wasn’t that great from my limited perspective as a technologist – although it was interesting to hear firsthand what Jeeves was doing after being let go and it was awesome to chat with Chris Pirillo, I didn’t see anything all that new and exciting there from a technology standpoint.

We were about to head to the IBDNetwork Web 2.0 event in San Francisco, and Al and I realized we didn’t bring any monitors for demos. If you saw us at the event, you know that we had the best screens, and, not arguably the best software. Luckily, New York City has one or two computer stores.  Here’s a photo of me carrying the screens from our hotel to the cab we took to the airport.  Why do I always get stuck lugging the equipment?

Last time we were in this situation, it was the night before C3 Expo.  We had all the signs and the materials ready and went to go duplicate them at Kinko’s, when we realized that it would cost us more to duplicate copies of our fliers than just to buy a color printer and pay for toner, etc.  So of course we bought the color printer and did the duplication ourselves.  The result was some great coverage on the local CBS morning news.Jayaudioblog

The demos at the Web 2.0 event went really well, and we met some very cool people, including Reg Cheramy, who was podcasting.

Anyway, to make a long story short, now I have a nice new monitor in my office.  Now I just have to take it out of the box!  And write an entry about Web 2.0.

-j

DR. JAY'S BLOG

  • Jay Budzik, Ph.D., is CTO of MediaRiver and coinventor of the contextual search technology powering ClickSurge. This blog talks about search, AI, and the media experiences that emerge from these basic technologies. More on Dr. Jay...

Your email address:


Powered by FeedBlitz

Recent Posts

August 2008

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
31            
Blog powered by TypePad