Slashdot Mirror
New Clustering Search Engine to battle Google
Sophrosyne writes "The New York Times is reporting a new search engine [free if DNA on file with Homeland Security] named "Clusty" is going to try and take Google head-on. The new search engine was developed by three former CMU computer scientists who formed the company Vivisimo. The search engine uses Overture for it's results but offers new features such as an encyclopedia search, clustered results, and a gossip search."
New Clustering Search Engine to battle Google
More like New Clustering Search Engine goes Beta. Let's wait until it's production stable before talking about who it's going to take down in a fist fight reminiscent of the Spock/Kirk battle in Amok Time.
Clusty by Vivisimo? Did I even spell that right? They need to consider naming things that people can:
A) pronounce
B) spell
C) are actual words or at least close to words that qualify for both A & B.
Clusty sounds like something you would call the fat cheerleader. It also will be often mispronounced as Klutsy, so it's a very bad name for a search engine (of all things).
The search engine uses Overture for it's results but offers new features such as an encyclopedia search, clustered results, and a gossip search.
This is a Microsoft tactic: add features to get market share, and it's an evil tactic because nothing new comes out of it, except bloat and bad karma. The fact this is based on Overature leads me to believe that it won't be able to take Google head-on at all. Clusty uses the Google interface but shows sponsored results first (evil), and displays 404 pages in the results. (FYI dteam was the first 3d design guild that is no longer)
I don't think they really have a hope of competing with Google. If it ain't broke don't fix it, so most people will just continue to use Google.
The dangers of knowledge trigger emotional distress in human beings.
But anyway, this does look interested.
I think there's your first clue for why your story was rejected.
The dangers of knowledge trigger emotional distress in human beings.
Instead of being able to search through just gossip, I would be more interested in being able to filter out all the gossip.
...but not one that reminds me of "Clippy".
So everybody is waiting for the next great search engine to come along and out-google Google, but it seems to me that they are looking in the completely wrong places.
All Clusty, A9 and the other more recent search engines seem to do is add more gimmicks to search results from yahoo and Google respectively. To some extent, this seems to be exactly what Google is doing recently as well: the searches are hardly getting beter, instead we can search news, search references (try define:), search printed text, do automatic conversions, etc etc.
But the truth is that not only are the searches at Google not getting better: they are getting worse. It seems like PageRank is more or less unused nowadays, and Google just uses easily manipulated things like searchterm in URL, searchterm in Title, how recently updated, to rank pages. I think anybody who uses Google to search for specific things must have observed that it works only a fraction of how well it did when it was new.
So what is going on here? Does everybody consider the basic searching a solved problem, and that we don't need to find pages better than google does? Or is a good search that cannot be manipulated really an intractable problem?
If I owned Google stock, I would really be wondering how many of all those thousands of PhD's at the Googleplex are working on this, and how many are writing gimmicks and elegant webmail applications. Or maybe one of them already proved that the problem can't be solved, and Google is just hoping to make as much money as possible before the secret comes out...
Under the heading "House" are the news items:
And under the heading "Record", are listed:
- As Reservoirs Recede, Fears of a Water Shortage Rise (Los Angeles Times)
- NASA Delays Plans to Fly Shuttle Soon (NY Times)
- San Jose State, Rice Set Scoring Record (AP)
This shows that just a clustering technique isn't enough; you need more context. Google (IMHO) does a better job of clustering their news results.Having said this, I wish Vivisimo all the luck. Google needs more competition; it is what will give us the Next Great Search Engine(tm).
Ob: I, for one, would like to welcome our new clustering overlords.. ;-)
The submitter had me all excited there for a minute or so, but unfortunately the "encyclopedia search" he mentions is simply searching the wikipedia.org site. Now don't get me wrong; there's absolutely nothing wrong with wikipedia, however it's already a web resource. You've been able to "encyclopedia search" Wikipedia for AGES by appending "site:wikipedia.org" into a google query.
/. about.
Bah.
Now if they'd done some sort of deal with Britanica to gain search access to its online library, THAT would be a resource worth posting to
Janie took my gun...
How many people actually jump on the "bandwagon" and switch search engines just because some one says it's "new and fresh"?
I gave a9 a try, I like the interface and some of the new features like the search history and the multiple search panes. But shortly after I found myself using google again. Even though a9 uses google, and the results are almost identical, I didn't find anything compelling enough to make me switch.
Does anyone else feel they might be missing some results if they were to use another search engine?
What must a new search engine provide to "steal" users from google?
Free iPods? Sure!
I have always considered Google's best point, is it's utter simplicity in design. Also, the name is easy to remember. Anyone who wants to up Google has to not only be MUCH better, but also have a good name and be as easy to use as google. Before, in the old days, each search engine produced sometime wildly different results. At the time, HotBot was the best search engine going, but they lost their steam and was ultimately "replaced" with google.
"Jeremy, you need to get to an internet cafe and cut and paste some appropriate sentiments about me from the world wide
Well, Google has got everyone beat in this regard. "Google" is probably the first thing a baby says (and hence I'm sure it is hardwired into our brains). The only thing that could beat "Google" would be "dada" or "burp". Any takers?
:-)
You joke, but a search engine named Dada would likely be well received for the name, and if it was a good system it could find a nice user base. I mean it has taken Google *years* to perfect its systems and they started with a good premise: do no evil. That was when all the search engines were cashing in on ads. A lot of people were turned off of the internet because of that, until Google came along. So it was purposeful, not evil, and light/easy to use.
My suggestion to anyone trying to take on Google is that they should do something else unless google becomes evil, and because power corrupts and absolute power corrupts absolutely -- it's just a matter of time before Google turns evil. Maybe not, though.
The dangers of knowledge trigger emotional distress in human beings.
First of all, "uses Overture results" strikes me as misleading. They have an agreement with Overture to share the proceeds from the sponsored links.
The results include MSN and Gigablast and Lycos. Basically, that means Yahoo's crawling plus Gigablast. Yahoo has ramped up their crawling since March, and is on a par with Google. They've been slow about passing all of it to MSN in a timely fashion, but by now MSN has most of it. I think Lycos, which also uses Yahoo's Inktomi, is about the same as MSN.
The clustering is the best of any search engine, meta or otherwise. You don't have to have JavaScript enabled, which is a big plus over the Vivisimo interface I remember from a year ago.
Finally, I was delighted to see that Clusty.com does not set a cookie unless you customize. Even the cookie for customization looked like it lacked a unique ID. I emailed Clusty and they confirmed for me that they have no plans for a unique ID in their cookie.
Google tracks you with a unique ID across all of their services, and saves everything it knows about you. Google's cookie expires in 2038.
Now I ask you, why do Slashdotters feel the need to dump on Clusty?
Now there's a first. Not even Google has ever directly supported Mozilla - the Google toolbar from Google is IE only. And this one now has a Mozilla search plugin link on the front page. Kudos.
Beware the psychokinetic mimes!
The basic concept of any kind of PageRank is flawed because it assumes a monotonic ordering of sites on some single scale (e.g., popularity as defiend by linkage). The problem with PageRank is not the use of links to assess popularity, but the presumption of a single scale.
The search of "Apple" illustrates this well. This search, like many is deeply ambiguous. It could refer to the computer company, to the fruit, to the record company, to New York City, to the singer (Fiona), or to Apple Valley (MN or CA). Even if the search engine knows that it refers to the computer company, it's still ambiguous. It could refer to the company (as an investment), the products (for purchase), or a question (as in technical support).
The point is that each of these ambiguous alternatives creates an independent cluster of hits. One cannot even rank hits within a cluster due to a hierarchy of ambiguity. Within the Apple computer cluster are distinct subclusters for computer purchase, investment evalaution, and technical support. Although one can create a ranking within each subsubsubsubcluster, it is impossible to construct a meanful rank for all hits across all clusters - the second hit for "purchasing an Apple computer laptop" is not comparable to the 2nd hit for "Apple Records".
Instead of a pagerank scheme that sorts the universe of hits the instant the user enters the search, search engines should be more interactive. The first page of hits would emphasize breadth -- displaying hits most representative of a broad range of alternative clusters. The UI would enable a "more like this"/"fewer like this" selection process that tells the search engine what the searcher is actually looking for. As the searcher selects hits, the subsequent pages might show popularity-ranked hits within the clusters that seem to interest the searcher.
Each hit and each page would serve a double-duty -- serving the searcher's need to get information from the internet, and answering the search engine's question about the needs of the searcher for that particular search. Until the search engine understands each searcher and each search, it cannot hope to rank the hits.
Two wrongs don't make a right, but three lefts do.
Finally, a search engine that correctly bubbles wikipedia above the spam clones (and read the reply to this post too). Google doesn't even show wikipedia at all on the first page, even if expanded. Kudos, you've won your first (?) customer!
A search engine that finds pages containing the words you typed which are *least* likely to relate to your actual underlying question. A google of the absurd, as it were.
This could be very, very difficult. How would you implement such a thing, from a technical standpoint?
The good and new comes from no quarter where it is looked for, and is always something different from what is expected.
Am I the only one who is fed up reading like "company A developed a new search engine which uses company B's search engine by adding revolutionary and world shaking features like thinking instead of you"...
If some are so revolutionary, then why are they using someone else's engine by adding some stuff most people most probably never find out what to use for. Doesn't A9 ring a bell for anyone, or does it.
I have an idea. Let's make a totally new and ground breaking search engine which will use Google's results, but hey, the main idea: let's have a different logo and paint the site pink !
Geez, I sometimes just can't stop wondering about all the freaky things that money can be earned from these days.
I am putting myself to the fullest possible use, which is all I can think that any conscious entity can ever hope to do.