Update: Forget Jeeves, ask Powerset
By Chris Gaylord | 06.29.08
Last month, I blogged about Powerset, the first Google competitor that really nails natural-language search. VentureBeat reports that Microsoft just purchased the Silicon Valley start-up at the rumored price of $100 million. To follow up on this news, here’s my introduction to Powerset from May, with some fresh analysis on what this buyout means.
Remember Ask Jeeves? The search engine branded itself as the web’s trusty maître d’. Type in your query – feel free to phrase it as a question – and Jeeves suggested where you could find an answer. But Jeeves turned off many users by directing them toward rather irrelevant websites. Since then, most search engines – including the Jeeves’ replacement, Ask.com – ignore all of those who, what, when, where, and whys. They just pluck out keywords.
Along comes Powerset. This startup website actually reads what you wrote. The search engine encourages you to write the way you speak, and then uses your phrasing to search entries in Wikipedia.
Type in “Who started Google?” and Powerset’s first response is portraits of Sergey Brin and Larry Page, Google’s co-founders. The labeled pictures link to their Wikipedia entries. If that wasn’t quite what you wanted, Powerset offers other links, just like any other search engine.
Try “When did Napoleon invade Russia?” and the search engine highlights “June 1812, Napoleon invaded Russia.” A search for the same phrase in Google comes up with several sites that have the correct answer, but you need to click through to find it.
Google has incorporated a few instant replies – answers that don’t require you to link to another page. The search engine field can convert measurements, calculate exchange rates, and answer “What is the population of Chile?”
But if you rephrase the question and ask Google “How many people are in Chile?” the search engine doesn’t answer “16,284,741,” as it did before. Powerset, on the other hand, answers both versions with the correct number.
The site is not perfect. For one, it only searches Wikipedia entries, so its pool of knowledge is incomplete and possibly inaccurate. The startup hopes to refine the algorithm and expand its sources, maybe one day allowing Powerset to search the entire web.
Update: Another strike against Powerset is that it’s really good at answering some questions and awful at others. I must have been lucky back in May because it answered almost all of my questions correctly the first time around. Now, as I play with it again, it’s batting about a .400. It seems happier with “who” questions rather than “when.” I bet this is because of its source – Wikipedia has comprehensive entries relating to people and places, and pretty meager posts on individual years. Let’s hope Microsoft helps the search engine improve its swing.
For now, though, Powerset is an early player in what’s called “semantic search” – that’s an ugly, technical term for teaching computers to understand natural language. Many futurists think that the “semantic Web” will be the next wave in Internet sites – a so-called Web 3.0. Expect a lot more of these natural-language options to come.
Update: Google has been a bit dismissive of semantic search, preferring (for now at least) its quick keyword approach. But this Microsoft news puts a lot of weight – and $100 million – behind the notion that web users want to ask questions to a search engine, not just feed it keyword clues. We have yet to see if Microsoft will keep the Powerset name or, more likely, integrate the technology into its Live Search. That site certainly needs some help. The company has fought a losing battle against Google and Yahoo for years now. Despite its best efforts and even cash incentives, Microsoft has not been able to distinguish itself. Offering a strong semantic search option is a good way to reboot the challenge.
But it had better get moving. Other semantic startups, such as Hakia and TextDigger, might not have the media attention that has been given to Powerset, but a sudden breakthrough or similar buyout from a competitor could quickly change the equation. Also, Google will continue to tinker with its search options. It already suggests that users try fill-in-the-blank search terms like “the parachute was invented by *” and let Google hunt for an answer to the asterisk.
[Via VentureBeat]
Also check out:
Microsoft: We’ll pay you to search
Search engines with more sparkle
No more Micro-hoo
Comments
2. Boyd Crow | 06.29.08
I entered the question “When did men first land on the Moon?” and the date of the moon landing was down around the fourth hit. It completely ignored the word “when” to give the first few results about “First Men in the Moon”, a book by H.G. Wells.
I find it to be no improvement over Wikipedia’s own search engine. What is the point of another search engine that “almost” responds to natural language search? How about putting all this money into tech support training so company respresentatives respond to natural language queries?
3. js | 06.29.08
isn’t it just a rumor? everyone is taking it as a done deal. Has it been confirmed by any source other than VentureBeat?
4. Mark Holton | 06.29.08
Powerset seems to be a step in the right direction for Semantic search. Once more companies get on board with this, it will cause providers to encode pages with proper RDF tags, and make webpages distinguish between synonyms by their meaning, not just their linkage. And vice versa, the more providers that start coding pages semantically, the more that mode of search will become important. Google has an amazing set of technologies and infrastructure, and they may be dismissive of Semantic web now (I actually doubt that assertion in the article), but that the Semantic web one area where the web’s capability can dramatically improve. REST plus the Semantic web have the potential to connect the web’s resources in powerful ways.
5. Ian | 06.29.08
In the 1980s AI researchers learned two important lessons. First that understanding natural language was really, really hard. And second that it really doesn’t matter.
Human beings are far more adaptive than computers. While chat-bots and natural language understanding have barely progressed in 30 years, humans can learn to frame their questions in search engine optimized ways in minutes. We are a looong way from having natural language search that is anywhere near as powerful as a human being with a small amount of understanding optimizing their search.
For those of us who lived through the first AI-bubble, it all seems very familiar.
Also the semantic web isn’t quite what the article suggests. As the commenter above says it is a way of encoding data so its meaning as well as its form is accessible to a computer. It doesn’t map onto natural language understanding directly.
6. Jim Bo | 06.29.08
Its important for search applications to emphasize the myriad subtleties of information classification to cater to different user characteristics. When these differences power search engines, we’ll get closer to perfect representation. Take a look at http://www.topodia.com = interesting approach
7. Larry | 06.29.08
What are you looking for? Just Powerset it and get wikipedia? Why not just go to Wikipedia?
And now that I have 2 desktops with Vista and Three desktops with XP the only questions I will ask Redmond is why Force Vista and why not let me just keep XP? I actually will not even ask them that as I know it is the $$$$$$$$$.
Gosh, with their fixation on Google and Yahoo, and knowing the MS history with IBM, maybe if they actually spent as much money and time working on a good replacement for XP they would then have more followers? Nahhhhhhhh …. Gates is out and Ballmer needs to follow him! MS looks more and more like IBM in the late 70,s and early 80’s.
I do not believe we will ever hear … if you want to find it … just MSN it or just LIVE seach it!
I just Googled Ubuntu, found WUBI and now have a great dual boot system that lets me totally avoid the MS BS!
9. matt | 06.29.08
Ian is right on track. I type “how many pecks to a bushel” into Powerset and nothing, I put the same thing into Google and it comes up with the correct answer. As far as Napoleon’s invasion of Russia I was inclined to type “napoleon’s invasion of russia” into Google and the first hit read: “Napoleon’s Invasion of Russia. In June of 1812, Napoleon began his fatal . . . “. Yahoo is similarly successful on both questions (but with a caveat or two). I was going to try the MS portal but I forget what it’s called and when I went to microsoft.com I didn’t find a link. So I guess Powerset hasn’t nailed search and neither has MS. Maybe MS should try to take one thing and do it really really well? But then again, branching out into everything under the sun is one strategy to increase shareholder profit, I recommend oil.
10. Vlad | 06.29.08
I wonder who uses them? I frankly don’t need this natural stuff. I appreciate if this gets integrated into Google but as a separate site doesn’t make sense.
I think people have learned to put relevant keywords to get specific answers from Google.
As someone has said, minor improvements in searches will not motivate people to switch search services. How is Google better then Live.com ? I don’t see a difference.
11. Amir | 06.30.08
This is step in the right direction! We have to move on… google was OK for couple years but we can do way batter than that! It saves a lot of time if you can form the question and get straight answer without going through all that garbage. Powerset will suck at the beginning just like google and yahoo did when they started…look at them now! If you have a right idea…sooner or later people will see value in their service and when they do, others will have to fallow or loose their market share!
Trackbacks/Pingbacks
Leave a Comment
We do not publish all comments, and we do not publish comments immediately. The comments feature is a forum to discuss the ideas in our stories. Constructive debate - even pointed disagreement - is welcome, but personal attacks on other commenters are not, and will not be published.
Tip: Do not write a novel. Keep it short. We will not publish lengthy comments. Come up with your own statements. This is not a place to cut and paste an email you received. If we recognize it as such, we won't post it.
Please do not post any comments that are commercial in nature or that violate copyrights.
Finally, we will not publish any comments that we regard as obscene, defamatory, or intended to incite violence.
1. Peter | 06.29.08
This came out a couple months ago on slashdot. I’m not impressed. Does about a tenth of what it claims too. It’s a fair search engine, but I think its major failing is that it claims to be natural language processing. That’s simply not the case, or if it is, its doing it very poorly. If this is the state of the art, I’d say we’re still at least a decade from real NLP