What is Digg?135 Comments
- balibones, on 05/13/2008, -5/+56I remain highly skeptical. Besides, I think Google does a pretty good job most of the time of extracting "meaning" from a web page to match my query. They are comparing their search engine with Google's "keyword based" search engine. It seems to me that anyone smart enough to build a "Google killer" would understand that Google is a "link based" search engine first and foremost.
- MaxMWood, on 05/13/2008, -2/+38Google Competitor not Google Killer. Lets get it right for once.
- Erect, on 05/13/2008, -1/+32I won't use it solely because they spell facts as 'factz' and tips as 'tipz'.
Whose genius idea was it to use z's ? They should be slapped. - isuisorisuaint, on 05/13/2008, -0/+25lol @ "google killer".
- playuhh, on 05/13/2008, -5/+30"For now, Powerset's index is very limited, consisting only of millions of pages from Wikipedia and Metaweb Technologies' Freebase, a Web-based structured database of information."
riiiiiiiiiiight... "For now we're pretty much whack! But just watch! We're on the come up! We're gunnin for that #1 spot google! Your ass is toast!
It's like when crappy rap artists talk ***** about the ones that have millions. Then you see em flop and the big kahunas are still making money and they just laugh. - drlha, on 05/13/2008, -1/+23Yes, but how many junk links does Google provide these days. For example, try looking up product reviews on Google and see how many actual reviews you find rather than just finding a bunch of crap like NexTag.
- .Steven, on 05/13/2008, -1/+19If Google buys it, how is it a Google killer? :P
"It's said that their understanding engine required a month to grok Wikipedia's 2.5M articles. The Web is currently at least 8,000 times as large." - Slashdot... - dgaspard, on 05/13/2008, -0/+16All google needs to do is to not point to this page in it's search engine. Then how will anyone know how to get to there site? Then the competition is gone. Problem solved.
- br0ck, on 05/13/2008, -0/+13http://www.customizegoogle.com/ is a Firefox extension that will permanently remove a custom list of spammy websites from your search results.
- mRIpX, on 05/13/2008, -0/+10please, no more of this *-killer of anything
- subtle, on 05/13/2008, -0/+10 Sorry, there were no results for your query: Why do you spell factz with a z?
- Corsix, on 05/13/2008, -0/+9No, they should be zlapped.
- surKaz, on 05/13/2008, -0/+7why is he getting dug down?... Just let him dream..
- munawarali2, on 05/13/2008, -0/+7Getting bought out by Google IS victory.
- surKaz, on 05/13/2008, -0/+6Hope-to-be/Wants-to-be Competitor... Just helping you and the submitter get it right..
- Jwoey, on 05/13/2008, -0/+7Who are you talking to? The submitter who copy pasted the article's headline? Or the author of the original article?
- tacojohn48, on 05/13/2008, -0/+6but wouldn't that be evil?
- rentmitchum, on 05/13/2008, -0/+5They probably hope google will buy them.
- Rotzooi, on 05/13/2008, -0/+4You might want to Powerset that.
- karebu, on 05/13/2008, -1/+5Rome was not built in one day.
good luck buddy, with catching up. - robbiemuffin, on 05/13/2008, -0/+5I tried those. I think you're nuts; they work well, very pretty too! Except I did "Caesar crossed the Rubicon" instead.
I decided to really stress the system: actors in World Trade Center (a less than popular movie with a lot of cross-reference in the name) -- no problems, great results again. The details are basically available directly on the second page in the results list from google's: actors "World Trade Center" -- so that's not amazing.
Anyone got a real test? - secrity, on 05/13/2008, -1/+6who is at&t's biggest competitor?
Came back with just a header and then a blank screen - Envark, on 05/13/2008, -2/+7Mmmm...delicious cyberpunk-style future where mega-corporations hire mercenaries to fight other mega-corporations.
I like it. - surKaz, on 05/13/2008, -1/+5Hey... they're trying to be hip... with all the hippity hop music and the slang and signs....
*sits in a corner and dozes off*... - rentmitchum, on 05/13/2008, -0/+4Yea.. yea it's just like that. Whack, yo.
- picklepete, on 05/13/2008, -0/+6Yeah, that has gotten progressively worse. I wish google had a feature, where if you were logged in, you had the ability to block domains from search results.
e.g., you search for "sony bravia review", and next to each search result there was an 'ignore results from this domain' button, or something similar. This would let you filter out the nextag & bizrate results at least. Maybe not the most elegant solution, but it would help. - siemersn, on 05/13/2008, -0/+3It's got a way to go and the 'google killer' moniker is entirely inappropriate. The semantic connections it makes across wikipedia are pretty weak in the few searches I tried. Everything is turned into 'noun - verb - noun' triplets.
- HonoredMule, on 05/13/2008, -0/+3Why should I have to "properly construct" a search to eliminate spam? The basic answer is that failure to do so results in seeing spam results. There was a time, however, when that was not the case. A reasonable understanding of the hurdles Google now faces would not stop me from switching to another engine that manages to produce better results, or do so more conveniently.
P.S. I use the -site: syntax regularly, for cases where a certain site is full of appropriate, relevant results, but is not where I want to go or the site I'm looking for. That, however, is quite a different matter from filtering out results that /no one/ actually wants or is not actually relevant anyway. - elucubra, on 05/13/2008, -3/+6Remember when Google was the ousider battling Altavista, Yahoo, et al...
Dont't write them off yet. let's try em' from time to time and see how they evolve.
I like competition, and I love underdogs that pull it off! - admdrew, on 05/13/2008, -0/+3Tried to do a comparable search on Google and Powerset of "what is the surface temperature of the sun":
Powerset:
http://www.powerset.com/explore/pset?q=what+is+the ...
Answer on 3rd result down:
"Solar furnace: The theoretical maximum is the 5778 degree Kelvin surface temperature of the sun , although the practical limit (due to atmospheric absorption and rapid heat transfer at high temperature differentials) is much lower."
Google (searched only English subdomain of Wikipedia):
http://www.google.com/search?q=site%3Aen.wikipedia ...
Answer on 1st result:
"[11] The Sun has a spectral class of G2V. G2 means that it has a surface temperature of approximately 5780 K, giving it a white color which, ..."
Searches on Powerset also take longer... the page loads quickly, but it takes a few seconds to "load articles." Though this may be changed eventually, we all know that Google is king for its relevant and *fast* results. I am not impressed with Powerset's current search offering (and yeah, the 'linkz' and 'tipz' thing bothers me a little). - theconnollykid, on 05/13/2008, -1/+4I had to say I chuckled somewhat at the irony when I saw that they are using Google Analytics on their site.
- dalittle, on 05/13/2008, -1/+4I tried the wikipedia searcher version and I was less than impressed. I will make another dabble at it, but I tried searching for "Myst Movie" and it only came up with the old Myst video game. Natural language search is the holy grail, but this one does not appear to be it. I smell VC funding press.
- klipseracer, on 05/13/2008, -0/+2Over 60% of all searches done online are done through Google.com Yes, that means over half of all questions asked online ask one entity.. Google.
- karebu, on 05/13/2008, -3/+6this is very sad.
go to http://www.powerset.com/ and see their examples
-------------------------------copied-----------------------------------
try these:
* paintings by Salvador Dali
* actors in Pulp Fiction
* when did Caesar cross the Rubicon
-------------------------------copied-----------------------------------
If you can't search by keywords, you're pretty much a goner.
For instance, the first query, on google I would have done
"Salvador Dali" paintings - dgaspard, on 05/13/2008, -1/+3I did a search for jboss vs glassfish. Google returns forum results of people debating the two application servers. powerset returns the jboss home page, nothing comparing and contrasting the two. Buried for not working. MSN search did better than this thing,
- secrity, on 05/13/2008, -1/+3what is the audio bandwidth of a telephone circuit?
Came back with 327 articles, none were correct. - secrity, on 05/13/2008, -0/+2Lots of big talk for something that right now is just an index to wikipedia and freebase. Even an alpha release of a web serch engine should index more than just two web sites. This is like somebody demonstrating a bear proof suit using Gentle Ben and a man dressed in a bear suit.
- MWeather, on 05/13/2008, -1/+3No, to grok, savvy?
- thomasprebble, on 05/13/2008, -1/+4"What is the Prime Minister of New Zealand?"
I hope they get better! - klaxxon, on 05/13/2008, -0/+2That particular search worked fine for me, returning relevant articles.
http://www.powerset.com/explore/pset?q=What+is+the ...
Am I missing something? - mrstorm, on 05/13/2008, -0/+3Natural language processing is a tricky beast. I do not think a single natural language processor can satisfy all users. Like a Bayesian spam filter - everyone needs their own for it to be of any use. Google's Page Rank is fundamentally different because it is NOT based on language - it is based on human behavior. More specifically, it is based on web author behavior. It cuts through the subjectivity of human languages. Pages are "good" because many people link to them.
If only there was an organization that had access to gobs of personal information about what people look at, search for, and write about in documents and email. Such an organization would be perfectly positioned to "train" natural language processors for each and every individual. Anyone know of an organization like that?
Here's what I do like about Powerset. It does seem to be a much better (faster) text search engine than the one built in to WikiMedia. Perhaps Wikipedia should adopt it as its primary search technology. - MWeather, on 05/13/2008, -0/+2It's pretty easy to tell what text is a meta tag.
- matschig, on 05/13/2008, -0/+2Same here
- Mysk, on 05/13/2008, -0/+2Wow dude, you need to lay off of the pixie dust... seriously, this HAS to be a very bad attempt at a joke.
- telepheedian, on 05/13/2008, -0/+2Wow, what rock have you been living under since 1998?
- secrity, on 05/13/2008, -0/+2In that case, then the article description was wrong. The article said that this is a beta test of the technology, not a sneak peek. I would expect a beta introduction of a web search search engine to actually have a sampling of Internet web pages; not just wikipedia (and maybe freebase).
- MxM111, on 05/13/2008, -0/+1The first line of powerset gives you 42.
- klipseracer, on 05/13/2008, -0/+1Exactly, if they became a threat, Google could just buy its stocks out and BAM. Its a Google compliment.
- EtherGnat, on 05/13/2008, -0/+1The problem now is that so many of the sites that DO have good reviews put up "review" pages for items they haven't actually reviewed. Video games and electronics are particularly bad about this. Search for a new product review and you'll get dozens of hits from well known sites, yet all you might get are "First Look" pages with nothing more than product specs and links to buy.
Maybe I could save a Google search with some common terms like "First Look" excluded to filter out some of the garbage, but it's an annoying problem. - klipseracer, on 05/13/2008, -0/+1Firefox != Google;
-
Show 51 - 100 of 135 discussions
Check out the new & improved 
© Digg Inc. 2009
— Content created and posted by Digg users is