Recently IDG News Service asked Google’s Marissa Mayer about the “perfect search engine.” Here was the question posed: “What is the perfect search engine? If you had a magic wand and could create it, what would it look like? What would it do?”

Mayer replied: “It would be a machine that could answer that question, really. It would be one that could understand speech, questions, phrases, what entities you’re talking about, concepts. It would be able to search all of the world’s information, [find] different ideas and concepts, and bring them back to you in a presentation that was really informative and coherent.”

What Mayer may have unknowingly described is Siri, a “virtual personal assistant” that uses artificial intelligence to determine user intent and then match data or applications that can fulfill that intent. The company will launch its iPhone application soon and already has a deal with a “tier one” US mobile carrier. The NY Times offers background on Siri and some of the technology behind the system:

SRI International’s software venture, called Siri, is more ambitious, in that it allows users to speak or write natural-language requests into the device (“Find me a place to eat dinner tonight with Karen, reserve a table and put it on our calendars.”), which will complete the task independently and inform you when it is done.

In terms of long-term predictions, Siri is actually an easy bet. Dag Kittlaus, the company’s chief executive, said one of the four major carriers would introduce the service early next year, and he said it would also be available as an iPhone app. But over the next two years the technology should be able to complete a wider range of tasks.

I’ve seen Siri in action and found it impressive. The system is not perfect but it brings users closer to transactions and fulfillment of their objectives — at least in a range of use cases — than can Google on mobile devices today. It uses a voice interface to receive queries. You can use the keyboard if necessary but that’s entirely secondary to the experience. 

The way one interacts with it is “conversational” and “transactional” rather than providing a verbal version of a conventional search query. 

I moderated a panel at the recent Open Mobile Summit in San Francisco called “new directions in navigation and search.” The panel, among others, featured Siri CEO Dag Kittlaus. What became clear during the panel is that we’re going to see lots of innovation and change in mobile search and that the present version of the experience could well be regarded as Jurassic in only a few years as the unique attributes of the device (e.g., the camera) become input mechanisms and search tools. Augmented reality is also a part of this, although in its present form it’s fairly undeveloped and limited.

And, as another example of how far things could develop away from the current “query box and blue links” search paradigm,  look at the video demo below of “SixthSense” a “wearable gestural interface”:

With a couple of Stanford graduates as founders, a new entrant to the online travel space officially launched its beta site today, WholeTravel.com.

Following a small but growing trend in the travel search space, Whole Travel uses what they call “Blue Sky Search”, or the ability for searchers to use open-ended experiential queries, rather than inputting specific location data, airport codes or exact dates for travel.

Searchers on WholeTravel.com are encouraged to type in eco-friendly activities (mountain trekking, wildlife tours, canoeing) or broad destination ideas such as “rainforest” or “desert”. With an editorial feature called “Whole Ranking”, results can be ordered by a scale of sustainability, based on factors such as on-site environmental and conservation efforts, renewable energy, contribution to the local economy, social and cultural experiences available to travelers.

The company has also started a non-profit arm, the Whole Travel Foundation, to increase sustainable travel practices, help promote and manage Whole Ranking, among other projects to help tourism businesses manage sustainability.

For the leisure traveler who considers sustainability, experience is the key motivator behind the purchase decision, so price sensitivity is less of a factor, though plenty of lower cost options (like tent camping or rustic cabins) can be found on Whole Travel, as well as resorts with high-end amenities.

Whole Travel has aggregated results from over 4,000 boutique hotels and eco-friendly resorts around the world, and will send visitors directly to their websites to book these unique accommodations (you won’t find many global hotel chains here), while they’ve partnered with Kayak for travel bookings.

In July, Jupiter Research published a report, Green Online Travelers: Assessing the Brand Impact of Travelers’ Environmental Concerns. Jupiter analyst Diane Clarkson had this to say about the potential growth of this market segment:

“We found that, among online green travelers, 7% are what we call “Dark Green”. These travelers have a favorable impression of travel companies with green policies and are actively green in their travel purchases (i.e. researching accommodation green policies, buying carbon offsets) etc.

WholeTravel will certainly appeal to this segment of Dark Green online travelers, for whom sustainability is a booking motivator. I believe WholeTravel will also have an appeal to travelers who, while they may not be in the Dark Green segment, are concerned about the environment and are looking for alternative accommodation options.”

Co-founder and CEO Matthew Davie said that in this market segment, up to 42% of consumers consider green aspects as part of their purchase decision. With travel being a $115 Billion dollar industry, this micro-segment has the opportunity to grow quickly.

After nearly two years in the making — and plenty of hype – Powerset has finally rolled out a "natural language" search engine. It’s not a Google killer. It’s barely a business model right now. But at least it’s something the world can finally play with, and under the hood, there’s lots of potential.

By the time you read this, the Powerset site should have changed into a tool that allows you search against material within Wikipedia. Why bother using Powerset rather than using Wikipedia’s own search tool or even Google set to look only within Wikipedia pages? The Powerset pitch is that you’ll get better results because Powerset’s technology has read and understood what every word within Wikipedia actually means.

An Understanding Engine, Not Natural Language Search

To understand that more, I beg that you forget you ever heard "natural language" being associated with Powerset. That’s not really describing what they do in comparison to regular search engines.

To explain, you have to understand that Google and the other major search engines are largely stupid. They don’t really understand the content on the pages that they "read." If they see the word "walk" in a sentence, they don’t know if walk is being used as a verb or a noun. In very general terms, they don’t even know that words are words. Words are more or less patterns to them — collections of letters — and when someone searches, they try to find the pages that have those patterns in them or in links to those pages.

That’s VERY simplified, OK? The major search engines DO have some smarts, some ability to know that walk is related to walking or that walk and run might be similar words. But this is largely done through statistical guessing, rather than comprehending what the individual words actually mean, especially in terms of their exact grammatical usage.

Powerset is different. It says that its technology reads and comprehends each word on a page. It looks at each sentence. It understand the words in each sentence and how they related to each other. It works out what that sentence really means, all the facts that are being presented. This means it knows what any page is really about.

In lieu of a better phrase, call it an "understanding engine." Maybe that’s not the right phrase, but natural language search isn’t it, either. Understanding engines at least highlights the uniqueness of Powerset — that’s because it actually understands what pages are about — it can extract facts from those pages plus comprehend how those facts, as well as those pages, relate to each other.

Wikipedia Discovery Tool

One of the chief uses for Powerset is employing it as a Wikipedia discovery or query refinement tool. To use the Powerset example they gave me during a briefing last week, consider a search for [henry viii]. What’s someone interested in in when they search on that topic, given Henry did a lot of things during his reign?

Over at Google, we get query refinement suggestions at the bottom of the page, like this:

At Yahoo

At Microsoft

Most of these are generated by looking at the relationships between those who have searched for one topic and then may have gone off and done another search. Yahoo has the most sophisticated of the pack (see Search Suggestions On Steroids: Yahoo Search Assist), but it still hasn’t actually "read" about Henry VIII and tried to group him into subtopics, in the way a human might.

That’s what Powerset tries. Here’s what you get in a search for Henry VIII:

Notice the tabs at the top, where it recognizes Henry VIII could refer to the person, the opera, the play, or even a television drama. OK, so not too amazing when you think about it. But look further to the "Factz" area. Here you can see that Powerset, after reading through Wikipedia, has figured out that Henry VIII "dissolved" things like monasteries or that he "granted" things like land. And yes, he "married" a few people.

There’s even more facts that can be found like this:

This is nice refinement. Running down the list, you can quickly scan the many facts that define Henry’s life. And from the list, with a click, you can drill in more about topics and jump right to particular pages within Wikipedia:

See how there’s a link to the Falmouth, Cornwall page? Powerset has seen that there’s something Henry VIII built mentioned on that page, Pendennis Castle. That’s not covered on the main Henry VIII page, but because Powerset has read both pages and understands what they are about, it can link the facts together.

Overkill For Now?

In short, the refinement is cool. What’s not to love about it? For one, it might be overkill. During the demo, Powerset made a big deal on how Powerset could build information from across various Wikipedia pages that isn’t written on any single one of them. For example, a search for [hulk hogan] brought this up:

See how those who Hulk Hogan has defeated are itemized? It’s nice — but do you really trust that all the defeats have been captured? I wouldn’t. I’d probably still go looking for an authoritative list that had been reviewed by a human. Moreover, I can get lists like that without great refinement. A search for hulk hogan victories on Google brings me to this nice page on About.com listing his world title victories.

In addition, while Powerset did a nice job of breaking down Henry VIII according to Wikipedia, Wikipedia’s human editors do a pretty nice job right in the opening paragraphs to the Henry VIII page:

Henry VIII (28 June 1491 – 28 January 1547) was King of England and Lord of Ireland, later King of Ireland and claimant to the Kingdom of France, from 21 April 1509 until his death. Henry was the second monarch of the House of Tudor, succeeding his father, Henry VII.

Henry VIII was a significant figure in the history of the English monarchy. Although in the first parts of his reign he energetically suppressed the Reformation of the Anglican Church, which had been building steam since John Wycliffe of the fourteenth century, he is more often known for his ecclesiastical struggles with Rome. These struggles ultimately led to him separating the Anglican Church from Roman authority, the Dissolution of the Monasteries, and establishing the English monarch as the Supreme Head of the Church of England. Although some claim he became a Protestant on his death-bed, he advocated Catholic ceremony and doctrine throughout his life; royal backing of the English Reformation was left to his heirs, Edward VI and Elizabeth I. Henry also oversaw the legal union of England and Wales (see Laws in Wales Acts 1535–1542). He is noted in popular culture for being married six times.

I suspect most people hitting Wikipedia are already going to find an opening paragraph like that, which does a pretty good job guiding them in refining their topics about Henry VIII and pointing them to facts.

That’s a problem for Powerset, which told me it hopes to  attract lots of those Wikipedia users to its own site, where they’ll be eventually shown ads alongside the content (ads aren’t present at launch).

Powerset was at pains to explain how popular Wikipedia is and what a well used resource it has become. Agreed — and plenty of those people wind up there because they’ve done searches at Google. About 70 percent of Wikipedia users come via search engines, according to Powerset itself. That’s a huge audience that is NOT going to magically be routed to Powerset instead. Yes, some know to go directly to Wikipedia. No doubt some of these users will hear of the new Powerset tool and go there. However, it will be a stunning achievement if these are more than a fraction of those who hit the main Wikipedia site.

Article Outlines

Powerset has another trick up its sleeve that might pull in the people. For any page you visit, there’s an "Article Outline" box that appears within it, like this:

It’s very slick. Select an item, and you’re jumped to the spot within the document related to it:

I think it’s self-evident that Powerset adds some nice value to Wikipedia. Indeed, everyone would probably be smart to go to it directly rather than Wikipedia itself. But as I’ve covered above, that’s not what I expect to happen.

Future In Site Search?

If Powerset fails to capture a wide audience, then what’s the way forward for it? One area is to provide better site-specific searching. Powerset’s technology can be applied to any set of documents, to make it easier for people to find what they are looking for within them. Site specific search allows those visiting a particular web site to look just within that site. That market, along with enterprise search (making intranets searchable) continues to grow. And the audience doing those types of search are likely more inclined to seek out refinement options and other exploratory tools than they are when performing general searches.

Powerset said this is a market they’re interested in, so perhaps we’ll see it grow in that area. But for those expecting it to produce Google-wealth, keep in mind that long-time and mature enterprise search player FAST sold for $1.2 billion earlier this year. Yes, that’s a huge amount of money, but it’s not the multibillions Yahoo was going to go for, and it’s much less than what Google’s valued at.

Speaking of Yahoo, it used to be the leading candidate in the past of who might acquire Powerset, especially given some close ties between the companies (Powerset has a number of former Yahoos on staff). Given Yahoo’s current troubles and unstable state, I wouldn’t expect much here.

Could a tie-up with a major player like Google or Microsoft happen? Sure. Aside from site search, the technology that allows machines to automatically comprehend what text documents are about ought to have other applications and be worth something. What those are and how much it is worth isn’t clear. Powerset’s been smart to snap up many licenses and patents around the technology that should make it attractive to a larger search player like Google or Microsoft to acquire. Within one of these organizations, I suspect more innovative things would come.

FYI, I wrote the above paragraph last Friday, before the rumors (see here on News.com and here on Techmeme) that Microsoft might want to buy it came out over the weekend. Actually, I started writing this article several months ago and in that, was covering how it might be an acquisition target. It’s a fairly obvious move to expect any of the majors to take a look, and when I talked to Powerset several months ago, I was given the impression that all the majors had taken a look.

Since then, of course, no one has acquired it — plus the company went through a management shake-up last year. It was already under fire for not getting a product out for so long. Add to these strikes as a potential Google killer the fact that it takes Powerset about a month to comprehend Wikipedia’s 2.5 million topic pages. In that time, many of those pages will have changed — thus needing to be reread again. Powerset’s impressive, but with the web having in excess of 20 BILLION constantly change pages, this is no overnight secret weapon that Microsoft might buy and employ to take the search lead.

Indeed, what Powerset says it has developed — along with patents locked up to protect it — is overkill for what’s needed today. It will be more useful probably five years from now, in ways we’re not even envisioning. For those players thinking long-term, which include both Google and Microsoft, sure — it might well make sense to buy.

By the way, the Powerset launch will no doubt inspire interest in another "natural language" search engine, Hakia. Someday I want to revisit Hakia and explain more about why I also dislike the term "natural language" being applied to it. In the meantime, you can read Vanessa Fox’s excellent article from last October on the service, Social Networking Through Search: Hakia Helps You Meet Others. And if you need a deflation of natural language hype, see The Google Challengers: 2008 Edition. In the section on Powerset, I summarize a long rant I did on the history and hype of natural language search.

For related discussion, see Techmeme.

Big management changes are going on at Powerset, which has received much attention for its potential in using natural language processing for search. Barney Pell, who has been CEO at Powerset, posted today in his blog that he is transitioning to CTO, that Steve Newcomb, who had been COO, is leaving the company, and that Ron Kaplan, who had been CTO and Chief Science Officer, is now solely Chief Science Officer. The company is currently looking for a CEO.

Several companies have been taking the natural language angle in creating the next generation search engine, but turning the potential into production has proven tricky. Powerset had hoped to launch their search engine this year but now thinks it could be as late as 2009 (although they are hoping for a mid-2008 launch).

Hakia, another player in the space, hopes to have their own brand of natural language processing fully powering their search engine next year.

Pell is optimistic about Powerset’s future. In his blog post, he says he’d love to talk to potential CEOs and closes by saying:

“The changes we are making now position us for a next phase that promises to be really exciting. We will bring our technology out in real products that users will enjoy and that will trigger changes across the entire ecosystem of search. I think the next year is going to be an amazing time for Powerset and I am as passionate as ever about Powerset, our technology, our team and our future.”

Hakia, a natural language search engine, has just added a new spin to search: social networking. Their new Meet Others feature lets you connect with others who are searching for the same things you are.

Since Hakia processes queries differently than old school search engines such as Google, you aren’t just matched up with people who typed in the exact query you did — you’re matched with a larger set of searchers that Hakia thinks are looking for the same things you are based on natural language processing. For instance, if you’re searching for “what drug treats a headache,” Hakia processes the semantic relationships between words and may deduce that someone searching for “what medicine relieves migraines” is a match. And that type of processing is the crux of how Hakia wants to differentiate itself.

I recently sat down with President and COO Melek Pulatkonak and CEO Dr. Riza Berkan to talk about what they’re doing in the search space and where they see things heading. More on how they’re tackling natural language processing below. First, a run down of what was launched today.

In addition to working on a completely new way of indexing and ranking the web behind the scenes, Hakia is also working on providing a unique search experience. In July, they launched the Hakia ScoopBar, which highlights the sections of the pages that are relevant for your query when you click through to them from the search results.

Today, with Meet Others, they hope to add a social networking component to search. The feature is entirely opt in. Once you do a search on Hakia, you’ll see a Meet Others icon above the search results. Click that to access a room designed for those doing similar searches. You can post a message and then provide details about how you want to be contacted (masked email or instant messaging via MSN or Skype). You can also contact others who have posted messages to the room. The freshest and most highly rated posts stay in the room longest. Older and less popular posts fall off as searchers make new posts. Hakia says they monitor abuse and have safeguards in place for spam (for instance, your post is authenticated through email).

What about privacy? Since the feature is opt-in, no one will see what you are searching for unless you decide to click the Meet Others icon and post a message. Even then, no registration is involved so your post isn’t associated with a username. And you decide what contact information you want to make available to others. If you choose email, Hakia masks it so anyone contacting you doesn’t see the address. (However, your IM details, if you choose to post them, are public.) You can also remove your post at any time, which removes any contact information you’ve made available.

Hakia likens this system a bit to craigslist.org. They want to bring people together. They use the example of someone looking for concert tickets. Someone else may post a message about tickets available for sale. Hakia can bring the buyer and seller together.

Here’s social networking search in action. Doing a search for “Looking for collectible pokemon cards” brings up results that include a Meet Others icon to the right of the search button.

Click that to see who has posted about that query.

Then post a message of your own or choose a contact option for someone who’s already posted.

As noted above, there’s no set time limit for how long a message remains in the room. It varies depending on how many messages get posted and how they get rated.

Is this an innovation in search or a recipe for disaster? How many people will find this valuable and how many will find it just plain creepy? In today’s climate of extreme social networking, Hakia just might be onto something, but the proof will be in the adoption.

Getting searchers comfortable with the notion of chatting with others about their searches isn’t Hakia’s only adoption obstacle. As I said in my review of one of Hakia’s CDs (they’re talented musicians in addition to computer scientists, who sing about zebras and finding your childhood on eBay for twenty five cents), regardless of how revolutionary their technology may be, the big challenge will be getting people to change the way they search because their technology isn’t at its best for the 2.8 word queries that Google has taught the world to type in. With those types of queries, Hakia performs just like any other search engine. Their differentiation comes with natural language processing, best used for longer queries that are typed more like the way people talk or write.

And what of this differentiation? Hakia provides interesting results now, but the jury’s still out on just how different and valuable what they are working on really is. They aren’t launching the search experience that’s powered by the core technology they are working on until sometime next year. They say what’s currently launched leverages the technologies they’re developing, so you can get a sense of what the final product will be like.

They say they are working on an entirely new infrastructure (different than what traditional search engines employ) called QDEXing (query detection and extraction). They search the web for concepts, rather than words, when satisfying a search. They point out that while the traditional search engines bring back good results most of the time, it’s impossible to know if pages that weren’t returned (because they have too few links to them, for instance) would have been more relevant for the query. By understanding the concepts on web pages rather than relying on things like external links and anchor text, they feel they can have a better sense of what page across the entire web is most useful to a searcher.

Traditional search engines use inverse indexing to catalog the words on a page. At the simplest level, when someone does a search, the engine looks through the index and finds the pages listed for those query words. Hakia, instead, uses QDEXing to determine what questions each page can answer. When someone does a search, Hakia finds that question and then returns the pages that answer it.

They use a number of scoring factors (such as linguistic and referential methods) to determine page quality. For instance, they say they can detect if a page was written by a person or was autogenerated. They only store pages that they feel meet their quality bar.

So, how will they get searchers to try it out? They may start with vertical databases that don’t do as well with traditional search technology. And they’ve cataloged the 700,000 most popular search results into galleries, which are algorithmically generated and have human review. Groups of hand-edited results for popular queries? Haven’t I heard this story before somewhere? Hakia says they’re different from sites like Mahalo and Wikipedia in that the results are algorithmically generated, so there’s less of an editorial component. They provide a well-balanced representation of query results — not just the most popular. And most importantly, the content is search results, not reference material. Everything about what Hakia does is about improving the search. What does a gallery look like? Well, I randomly chose a query — “Buffy the Vampire Slayer”. No reason. Hakia returns a gallery page with things like headlines, television profile, the channel it’s on, pictures, and fan sites. Hakia algorithmically generates the categories based on overall semantic processing of pages about the topic.

Of course, there are other players on the natural language processing bandwagon, with Powerset one of the most hyped of the bunch. Will Hakia’s approach of providing a unique user experience set them apart? Well, the Buffy page is kind of cool. Whether or not searchers can get comfortable with a new search experience and different way of querying remains to be seen.

Both VentureBeat and the New York Times are out today with an update on Powerset, which has won the rights to natural language search technology from Xerox PARC that it hopes will make it into the new Google, when Powerset finally launches. VentureBeat also gives us news of at least one person from Yahoo now working there – Tim Converse — who left Yahoo in December. I know others from Yahoo are also at Powerset, as well.

Last time we had news about Powerset, I did a big long rant – Hello Natural Language Search, My Old Over-Hyped Search Friend — about the hype and history often associated with natural language search. VentureBeat mentions that rant, as does TechCrunch today in its post, Powerset Hype To Boiling Point. So I felt I should give an update.

Unlike Matt Marshall over at VentureBeat, I’ve yet to see Powerset (right now, you can only see it within the Powerset offices, and I’m in England, not Silicon Valley). I have been in touch with CEO Barney Pell, and I’m looking forward to using a working demo in the future.

I also have to stress that not having seen the tool firsthand, it could be that I’ll be blown away and as impressed as some others are. Then again, right now I still feel that it’s more hype than actual Google threat.

Technically, I’m "off" from writing on Fridays. So for now, I’ll just share what I sent over to both VentureBeat and Barney last night:

Frankly, I can use Hakia right now and have been very impressed with much of what they deliver, though not all of that is due to the natural language stuff. I have an article I’m working up on it to talk about how they are better for some queries, in some respects.

I knew about some of the Yahoo people going over, including Tim. Interesting, certainly gives more to the Powerset story to consider, but I’m still firmly in the prove it category by actually rolling it out.

who acquired ibm
[a search cited as Powerset doing better than Google in Matt's story]

OK, I see Surfaid, Lenovo listed, so that’s two "IBM" things acquired. I’m not sure who searches for that [particular phrase] to find out who acquired particular IBM units. That type of query to me feels like something you were suggested to try, and if so, then usually it’s suggested because it’s known that query will produce something good. Go search for travel, cars and casino and compare those to Google and see if you’re still impressed. Those are real searches.

Also, what’s the size of the index Powerset is searching against? Barney?

Google has in excess of 10 billion documents from across the entire web, and that’s a low estimate. If Powerset is hitting fewer documents — and especially a subset of say business documents — just that filtering along might give you better results.

So later, when you ask Powerset: "What do liberal democrats say about healthcare policy?" And that sounds really compelling. But guess what? On many pages that say liberal democrats, you’ll also find Hillary Clinton also mentioned.

• What do liberal democrats say about healthcare policy: 1.4 million matches
• What do liberal democrats say about healthcare policy hillary clinton: 900,000 matches

Maybe that special document I needed was in the 500 million that Google’s not picking up, because it’s not making the association. Chances are, it’s not.

"Powerset showed it can answer more complex questions, such as “Who did IBM acquire in 1996?�? Here, Google completely breaks down
[quote from the story]

Breaks down because what did Powerset show? I mean, a screenshot? A list of what was there?

FYI, if I do these:

• ibm acquisitions 1996
• ibm acquisitions

Both bring me to Wikipedia first [via Google] which answers the question in short order and is anything other than a failure.

Now ask yourself. If you’re looking for an IBM acquisition in 1996, don’t you think you’re likely to wonder about a particular one? Or if you’re after things they’ve acquired, are you that likely to put in a particular year?

Sure, some refinement suggestions would help. But you’re picking out one query and saying Google breaks down when that might not even be a query someone would do.

Clearly, Powerset faces challenges. Even if its technology does prove to be useful, it isn’t clear how long it will keep its lead in the face of an onslaught from Google. Another challenge is changing peoples’ search behavior, which is used to keyword searches.
[quote from the story]

What lead? With respect, it doesn’t have any lead. It has some experts, a product pitch and a product a few can play with and only internally. When it launches, then it has a serious challenge to take on even Ask, much less Google. I’m not saying it can’t be done — but wow, you’ve already got it winning while it hasn’t even entered the race. Not only does it have to attract traffic, but it needs to prove it can scale to handle that traffic and the processing. Will it stand up to doing all this natural language processing (which often will be unnecessary) in the face of a lot of traffic.

NOTE: Matt also replied to me:

By lead, I actually meant lead in natural language search. Clearly they don’t have a lead in search. (though, even in natural language, I suppose you could say that is speculation, given that [Google] tells me they’re got people working on this).