Quintura Search Engine gets backing from early Skype investor
Written by Alex Iskold and edited by Richard MacManus.
Russia-based Quintura Search Engine has received substantial backing from Mangrove Capital Partners of Luxembourg, an early-stage investor in Skype, ABRT Venture Fund and the partners of OpenView Venture Partners of Boston. This is an interesting deal for several reasons. The first is the Skype connection. Another one is that Quintura is the first Internet Russian company to receive investment from a Western VC. And then, of course, the product itself is very innovative.
Quintura is a visual search engine and it claims to apply the natural way that people think about concepts, to search. One of the most difficult problems that modern search engines are trying to solve is resolving ambiguity. Quintura solves it by presenting the user with a cloud of concepts relevant to the search. For example, when we searched for readwriteweb we got the cloud shown in the picture above. This type of cloud helps to refine results. And when a user clicks on a word, the search results below re-arrange accordingly.
The current version of Quintura supports just generic and image search, but it is likely that more will be coming soon. Users are also able to save and share the results that they find relevant. The results window could use some polishing, but the cloud navigation definitely compensates for any shortcomings. It is clear, as the company claims, that some very sophisticated neuron networks and clustering on the back end are powering this cloud.
Yakov Sadchikov, co-founder and CEO of Quintura, said in the press release:
"We plan [on] using the venture capital to take the Web by storm and reach millions of web users globally when launching the Quintura visual find engines for various user communities."
We think this is very ambitious, given that search is a tough and crowded space. But with a bit of polishing, this one may have a chance of breaking through. Could it be that Mangrove have found another big winner? Only time will tell. For now give Quintura a try and tell us what you think.
Share1 TrackBacks
TrackBack URL for this entry: http://www.readwriteweb.com/cgi-bin/mt/mt-tb.cgi/2866
Comments
Subscribe to comments for this post OR Subscribe to comments for all ReadWriteWeb posts
Seeing results in a very limited 400px space?? No, thank you.. I don't like these kind of "fancy" web 2.0 ideas in search engines. Search is a utility and should be kept as fast and simple as possible.
I love the innovations SearchMash is bringing but that's all.. As for clustering search results, this is not a new idea.. Clusterix and Ask already does something like that.
The interface needs work-- the response rate for mousing over a word and seeing words clustered around it was so slow that it took me a while to figure out what was going on. Also didn't find it intuitive to click a keyword to get rid of it. For some reason, image search gives much faster feedback for me, and I do very much enjoy the instant visual feedback there.
This seems useful in suggesting a few keywords that will help refine a search, but the cloud doesn't seem to do much for it. When I searched read write web I looked around a while to try and find "blog." Perhaps its just that the cloud-creation algorithm needs a better method of deciding how to organize the keywords in space. The keywords most often combined with that word or that give most results should be the closest and biggest.
The info aesthetics blog has some related search interfaces linked to from the latest entry on Riya - http://infosthetics.com/archives/2006/11/like_visual_search.html
I found the UI intuitive, responsive and intelligent. For me, Quintura is on to something, I definitely enjoy drilling-down into a tagcloud, it seems efficient. The diversity visual between the cloud on top and the results listed below is compelling. On the other hand, I was a bit disappointed by the search results. The interest is to be able to see a complete range of semantic search possibilities from a search query into which one can drill. But the results didn't seemed geared to this: ie. I searched the word "space" (anticipating "architectural" or "virtual" "space" and got nowhere near these possibilities, coming up with the same old, same old "nasa" "station" etc... Good post, nice find.
Alex, my dear friend, http://startupsquad.com/2006/11/03/quintura-visual-search-engine-startup-gets-funding/
If you're interested in concept search engines like Quintura, you might like my new search engine, Zeedex.
Like Quintura, Zeedex gives you suggestions to refine your search. You can then click on any suggestion to add it to your search.
Zeedex has two special features:
1. The suggestions are categorized in lists (e.g. a search for dogs returns a list of dog breeds, dog toys, dog diseases, etc.).
2. Anyone can add to or edit a list, like wikipedia.
Please check it out, and email me your throughts.
russell@adamm.net
I've tryed tiger and got these results: http://www.quintura.com/?request=tiger&searchvia=0&engineoverride=0
Then, i've tryed the same search on Ask and got these other ones:
http://www.ask.com/web?q=tiger&qsrc=0&o=312&l=dir&sugreqs=3
So far, Ask seems to be more effective in helping you to define you search field.
Ciao
Nicola
Isn't it similar in some ways to www.cartoo.fr ? And this search engine is around for more than 5 years.
Errata on the above message. It is www.kartoo.fr
Not performing at all.
Not a real innovation, there are other mapping search engines...
Certainly a way to fund the Russian Mafia!!