Top-of-mind thoughts on Microsoft and Powerset

There’s a reason I love emerging technology so much: over the course of one hour, the entire landscape can be turned on its head. The rumor out of VentureBeat this afternoon, that Microsoft will acquire Powerset for $100 million next month, has produced the predictable memes: Microsoft is desperate after the Yahoo debacle; Powerset overhyped itself to bankruptcy and needs a bailout; Powerset only searches Wikipedia and we like Google just fine, thanks. While neither party will confirm the rumors, it now seems likely that something significant will happen in the semantic sector over the next couple of months. Having analyzed Powerset and semantic search extensively, I think we should keep a couple of key points in mind beyond the arguments over valuation and hype machines. Continue reading

Sites I’m Loving This Week

I’ve fallen off the blog hamster wheel in recent weeks, due to travel and screening companies for DEMOfall. My calendar gets a bit ridiculous around the same time twice a year, as I spend entire days on the phone hearing about new companies. Mind you, I’m not complaining. Even when some days morph into one continuous conference call, it’s still one of the best jobs around. Paradoxically, all this activity precludes my favorite job: telling everyone about the new toys I’m using. So if you don’t mind a laundry list, here’s what’s been on my radar lately. Continue reading

Geekery Fiefdoms at SemTech

For all my stewing about presenting an effective panel here at SemTech, I think we did it in spades this morning. I’m biased of course but if the amount of active, engaged audience members and lively conversation following the panel was any indication, Taking Semantic Technology to the Masses was a success. Thomas Tague, Josh Dilworth, Mark Johnson and I had an excellent discussion about the mess the semantics space is currently in, marketing-wise, and how to dig it out and shine it up for mass consumers. We spent the first 25 minutes parsing the problem – an indication of just how deeply semantics geeks can gaze at their navels – and about 20 more minutes discussing possible solutions.

Thomas coined a term I’m stealing that sums up the semantics space perfectly: geekery fiefdom. It’s a great description of a sector that is striving to achieve traction in the consumer space, but continues to pepper its messaging with semantic buzzwords and discussions of the plumbing behind it all. As Thomas quoted one of his customers in the financial sector, “If you have to explain it, I don’t want it.”

We came to a couple of good conclusions worth mentioning:

1) Companies in the semantic space need to take a portion of their impressive brainpower and turn it toward marketing. With literal rocket scientists on the benches, finding innovative, well-packaged messages around a product and company philosophy should be a piece of cake.

2)UI, UI, UI. Mark mentioned this several times and he should know; Powerset has one of the best out there right now. Once you’ve parsed out the complex algorithms of your semantics company, spend some time on a great design. An easy-to-use, intuitive interface can vault a product to the head of the pack.

3) Play nice and share. (I’m reminded of that annoying book/poster from the early 90s – Everything I Need to Know I Learned in Kindergarten.) It’s simple but true. If semantics companies were more open to partnering with each other, the resulting applications would without a doubt take this industry to the next level. The close-to-the-vest attitude is understandable in semantics, as some very sophisticated and complex platforms and algorithms are at stake, but I think we’ve reached the point where it’s time to open up a little.

Everyone seemed to agree, including members of the audience, that semantics is poised to graduate; that it’s time to dust off this fiefdom and take it out into the countryside among real users. When and how that will happen is still undecided but I’d bet on later this year or early next.

That’s it for the moment from SemTech. I’m huddling with Hakia in a bit and can’t wait to hear their news, then it’s time to concentrate on the French Tech Tour for the next 12 hours. More tomorrow…

Search redefined

Anyone with even a remote interest in the semantic space has likely experienced the same roller coaster I have regarding Powerset. When I first spoke with Barney Pell over a year ago, the semantic tech sector was an entirely different landscape. I was intrigued by my conversation with Barney and the short demo I saw of Powerset-enabled search. How nifty that the engine knows what I mean by “who did IBM acquire”! But as months went by, we didn’t hear much from Powerset, save a seemingly incongruous Labs announcement. And we heard much from other players in the space. The focus of the semantics community moved away from search to organization – making users’ Internet activity easier to manage – and answering the question of how to take semantics to the masses. Frankly, I had dismissed Powerset as an early mover in the space that had run out of steam. Boy was I wrong.

Powerset’s introduction today of its new Wikipedia search, which also integrates data from Freebase, could have a significant impact on the tech market overall, in that it changes the rules of the search game. Users who experience the incredibly deep, interactive, and intuitive nature of the Powerset search will be even more frustrated with the standard string of result pages delivered by traditional keyword search. Once you’ve dug into the meat of a Wikipedia article with just a couple of clicks, zeroing in on precise actions and entities and going directly to their citations in the article, paging through flat hyperlinks just ain’t going to cut it.

Powerset’s changing of the rules is evidenced by one key statement made by the company: a page of search results, no matter how targeted, is just the beginning of the effort required by the user. Once you’ve found relevant links, you still have to click through to new pages and scour the text for usable information. Powerset’s new way of searching attempts to do some of that work for you; with the scouring and drilling down already complete, you arrive at what you need much quicker.

The Outline feature of the Powerset search is a real gem and I expect will set a new standard for UI in search technology. Having a constant window beside the text as you browse provides an incredibly simple way to jump back and forth between concepts and facts. It could make the browser’s back button obsolete.

What I don’t love about the new search is that it’s currently only on Wikipedia. There are many searches I typed in that can’t take advantage of all this whiz-bang semantic technology. More nebulous concepts aren’t Wikipedia’s strong suit, so Powerset only returns standard results. Example: “Can Hillary win the democratic nomination” returned relevant results but no Wikipedia entry to plumb. So my big “if” with this announcement is whether Powerset can pursue a successful content partnership strategy. If the right publishers, and enough of them, integrate Powerset search into their sites, the long-anticipated threat to Google could finally take shape. No matter the long-term outcome, though, Powerset has raised the bar for search interaction and usability.

A Second Opinion

I wanted to dash off a quick note and point everyone to an interesting conversation that happened here on The Guidewire. I’ve written about Songkick a couple of times recently, giving the service praise for its focus on semantic-based recommendation for consumers. In my most recent post, I specifically raved about its new capability of recommending concerts to users based on their music preferences. My friend Shellee, very much a tech outsider and a live music fanatic, gave Songkick a spin and wasn’t as happy with the results. She said so in the post’s comments and, after a request from me, the company responded to her in kind.

I say this not to needle Songkick, who posted an excellent, well-reasoned reply to Shellee, but to again stress the importance of the mass consumer, a theme we return to repeatedly on this blog. Rave reviews from The 250 are great in the short run, but end-user stress-tests are the only reviews that truly matter in the end. It’s a good lesson and ego-check as we tech insiders continue to debate our role in product analysis.

The Other End of Semantics

I wrote about Songkick last week, praising its focus on technology for the mass consumer and referenced an impending announcement. That announcement came yesterday, with the launch of a music recommendation service. Put three bands you like into the system and it returns recommendations of area concerts you might enjoy. Simple but brilliant. It works pretty nicely, too. I typed in two current obsessions – The National and Vampire Weekend – and one mainstay, PJ Harvey, and it returned Radiohead and The Cure. Two concerts I’m actually interested in seeing and will try to get tickets to. It should be said that I did stump the engine by throwing U2 in once. But perhaps it’s trying to tell me I need to update my music library.

What I like about Songkick, as previously mentioned, is that its creators aren’t interested in parsing the ins and outs of the technology. They instead want to spread their love of music through enabling technologies. I called it “music semantics” and, though the pundits in the semantic realm may take issue with that label, it’s time we embraced apps that are less wonky in their approach and focus. While Twine, Hakia, and MetaWeb are laboring in the code mines, working to build what will be the framework for the semantic Web, companies like Songkick are out in the market, showing consumers real-world applications of semantics. It’s vitally important all such players are represented, in order for semantics to develop fully and organically.

Twine is Lifeline for Semantic Tech

It’s always fun to watch the evolution of a great idea. I first talked with Radar Networks, Twine’s creator, a year ago, in January 2007. In fact, Radar’s was the inaugural profile of The Guidewire Report, kicking off our in-depth look at up-and-coming companies with a bang. Unable to speak specifically to the Twine application (which wasn’t even named at that point), I mixed my praise for the idea with a bit of skepticism.

A Web that learns from its users and manages the infinite amount of knowledge available, in a unified Web-based environment, is without question where technology must head in the next few years….Convincing set-in-their ways consumers that a new form of communication, collaboration and information management is needed is perhaps the biggest challenge. Educating users on the semantic Web and why it should matter to them may require more effort than is reasonable.

It’s surprising that not much has changed in a year. The landscape in which Twine launches is just as undefined, if not more so. The chatter is rising to louder levels, as I mentioned in a recent post on semantics, making it that much harder to clearly define this important but hazy sector. Even worse, any application tied to the semantics label these days carries a heavy weight. So many are touting the life-changing aspects of semantics that disappointment is all but inevitable. By the time you’ve read dozens of articles on the brilliance of semantic apps, you half expect these programs to answer your email and write your blog posts.

It’s into this environment that Twine is finally opening up more – to the press, as of this writing – with 30,000 people on its waitlist and arguably an entire market sector watching. So…. what’s the verdict? Continue reading

Super Search

As I’ve mentioned before, I’m one of the many obsessed with the semantic Web. It’s a catch-all term for anything, really, that makes Internet content more intelligent. But recent favorites of mine involve connections; tying together seemingly disparate entities online can produce insight heretofore impossible. Two companies here at DEMO have developed incredibly sophisticated algorithms that have changed the way I work. If you do any amount of research online – and who doesn’t these days – run, don’t walk to Silobreaker and Jodange.

I’ve been addicted to Silobreaker for several months now, so was delighted to see new features and a more intuitive design to the site. A current affairs search engine that combines context extraction and relational analysis, Silobreaker provides a 360-degree view on news events, people and places. I know, that sounds somewhat buzzy; you’ll just have to check it out for yourself. I find the network search especially helpful, which offers unprecedented insight into relationships between people and topics around a specific event. I spend a huge amount of time researching and have never stuck with a site so faithfully. Almost without exception, it delivers precisely what I’m looking for every time.

Jodange, which has understandably had a few meetings of like minds with Silobreaker, provides invaluable insight of its own. While Silobreaker pulls salient quotes as part of its entity analysis, Jodange focuses entirely on opinions. Its ‘sentiment analysis’ engine, called TOMS (Top of Mind Service), mines and indexes opinions across the Internet, allowing for a distinct perspective on significant topics. How negative was Hillary Clinton the week of November 17? What is the oil industry’s view of the recent rate cuts by the Fed? Answering questions this specific was previously impossible without hours of research. You just can’t find this stuff on Google. Period.

Powered by ScribeFire.