Comments

Subscribe to comments for this post OR Subscribe to comments for all ReadWriteWeb posts

1. Good one ..Yes search has long been monopolized by Google..Web2.0 stands a good chance now.
   I go with sites like Delicious[people's voice is the best one..]

   Posted by: Casey | January 5, 2007 4:11 AM





2. I really doubt how web2.0 companies are going to beat Google.Lets face it Google does provide the best search results among all search engines at the moment.Best in the sense that its searches are better than most of the other search engines.Web2.0 is yet to become integrateed among the mass users.Web2.0 companies has to come up the value chain and deliver customized results that takes into account the preference of the user.Except Myspace.com I have not seem any web2.o companies becoming accepted mainstream . I have no doubts that the day is gonna come when Google will be challenged but not till the nature of the way the game is played changes. People come to Google to search and until people stop searching I dont think any of the above could actually challenge Google . They should be playing to Google weakness and not its strength( Clustering,Vertical searches,AI are all playing into Googles core strength )

   Posted by: Ron Jones | January 5, 2007 4:30 AM





3. gI agree with Ron Jones regarding the adopting mass. not agreeing about the quality of search results. think 10 years from now, will search results be the same? things are changing. google made as accustomed to search results that are not personalized in the sense of not 'speaking the language of the user' not addressing his expertise and specifics. eventually it will. it is just a question of time. whether it will be Google or the uprising Collarity, that for the people to decide. by the way, Collarity results are far better for many cases than Google's after applying their personality mode. people will continue searching for answers and relevant information, that will probably remain part of user- web interaction. the evolution will surly come from knowing better the user and his needs.

   Posted by: yavin | January 5, 2007 6:41 AM





4. I think that Quintura with 29% of votes is winning the race. It is AI (23%) plus visualization (6%). I've just checked their corporate web-site to see if they are based on AI. Yes, they are.

   Posted by: John | January 5, 2007 7:50 AM





5. 
   google has built a significant brand, tough to beat. But, the market needs social search as a way for users to contribute. Sproose.com offers a user moderated search engine...it is more important that users have creative choices, the soft drink market would be boring if there was only Coke!

   sproose- ing

   Posted by: Bob Pack | January 5, 2007 8:50 AM





6. I go with people-powered. Of course I'm running a people-powered search site, so I may be biased.

   Posted by: Adam Jusko | January 5, 2007 9:28 AM





7. It doesn't matter what technology being used inside. As a user I need a search engine that gives what I really wanted. No search engine can do that at this moment. Google is the best one closest based on relevance of your keywords. And Google has a knack of innovation and constant improvement. So I bet on Google my 99 cents. But I love Quintura as well for Visual interface. But if Visual interface gives enough advantage to Quintura to beat Google, I guess Google will develop a much better Visual interface long before.

   Having said that, no body CAN beat Google as far as Search Results/Technology are concerned. I believe, if there is a better way to search, Google will leverage it too. What prevents Google from doing so?

   But if there is a unique business model these new startups can come up with, then there is real chance to attack Google. So advantage could be by virtue of business model not definitely on technology.

   Traditional wisdom says, 'Slow and Steady Wins the Race'. I would add, if only the opponents are sleeping.

   Disclaimer: I am big fan of Google and love Google Products by virtue of Elegance, Efficiency and constant improvements.

   Posted by: Murali | January 5, 2007 10:57 AM





8. This poll is really interesting, but it has some challenging semantics associated with the categories. In my experience, AI is a term few understand (not sure I do) or agree on definition-wise. Nevermind… it’s a cool poll!

   Consider the people-powered challenges that have been overcome by folks at companies like Digg and Wikipedia. Five years ago if you said, ‚ÄúI‚Äôm going to let users-at-large control headlines and content‚Ä?, it was easy to assume collective intelligence would be trumped by chaos and self interest. Today, while neither is perfect (still battling the spamming horde), each has developed a valuable alternative mechanism (not based on link topology) for serving useful content.

   For Collarity, we believe that better filters (personalization) will enable better recommendations (community-driven search results). Paradoxically, we also believe that once you have a well defined personal filter, you shouldn‚Äôt have to go looking for stuff as much ‚Äì useful and interesting information should arrive where, when, and how you designate because other people (the ‚Äúright‚Ä? people) have found it.

   I don‚Äôt know about you, but I do less searching today because of my trusted feeds (‚Äúhere‚Äôs an interesting new development‚Ä?) and my favorite people-processed news/ reference sites. It‚Äôs really about where attention migrates in the future and the advertising dollars associated with those new business models.

   Posted by: Rob Rustad | January 5, 2007 1:31 PM





9. The results don't have to be substantially better than existing search engines. They actually only have to be better enough to persuade people to switch some or all of their queries. In some cases this will be very tough because Google will do an exceptional job responding to queries.

   There are really three main elements where natural language search can provide a clear distinction of existing modes:

   1. The firs tis in query disambiguation: as a commenter points out, Google already handles long queries and does a better job of post-processing than it used to (stemming, handling singular vs plural, misspelling). But it isn't necessarily capturing deep semantic information from its queries (something we used to do/infer at Albert.com)

   2. The second is in the indexing. Google has a powerful index based around noun-phrases and link structures. To really amp up the value of an NLP search engine, you need to be able to index more semantic information--this has been computationally and storage expensive. But query processing can only take you so far if your index isn't really rich and detailed.

   3. The third is whether your semantic processing capabilities allow you to drive a better revenue yield from search queries. Google's secret sauce has really to be not just the attractive economics of its search offering (as a webmaster, offering search isn't just about improving the UX of your site but of improving its basic business economics) but its also that its search economics are orders of magnitude better than those offered by Yahoo, etc. However, if natural language processing can change the equation for search economics . . .

   best
   Azeem

   Disclosure; I am an investor in Powerset.

   Posted by: azeem | January 5, 2007 4:57 PM





10. I don't know if it's important, but it looks like you can vote multiple times with Polldaddy. Just navigate away from the page with the poll applet, go back, and you can vote again.

    Posted by: Juha | January 5, 2007 5:09 PM





11. Collrity is a powerful and easy to use SE

    Posted by: Ilan | January 6, 2007 1:49 PM





12. I start using collarity and it performs great. it quickly suggests the correct words to place in the search phrase.

    Posted by: Avi | January 6, 2007 2:26 PM





13. I personally think interative Q&A-like website is a good complement to search engines. People can ask questions when they cannot find what they want. This is proved by why Yahoo Answers, Baidu Zhidao have attracted a lot of attentions. However, Google Answers was a failure. I think there are two main reasons why Google Answers failed.
    1. Their business model was too strict, 2-200$ per question. I think if they can allow 1 or even zero (as flexible as BuyAns.com), more people would like to try.
    2. Google Answers is not true Web 2.0. Only 250/500 search experts can answer (and lock the question). No opportunity is given to general users and other experts and there is no real competition. Maybe many general users can provide better answers.

    I don't think ChaCha.com can survive even though they got 6M VC recently. People's demand and new web pages will keep increasing and their method cannot afford paying to their expert always. This is why I like BuyAns.com like Q&A website, which rely on users, not the site itself to pay for their requests. In the meantime, the website can automatically annotate the questions/answers since questions/answers patterns are used for asking. Meanwhile, the patterns can be created by general users and shared with others. This is the sense of Web 3.0.

    Posted by: Liu Wenyin | January 7, 2007 12:45 AM





14. Do you really think Google can't implement following technologies in their existing search technology?

    ...1. AI 2. Ppl Powered 3. Vertical 4. Personalized 5. Clustering 6. Social 7. Visualization 8. Previews...

    Sure they can, but as expected from a smart company like Google they are playing a wait and watch game.

    1. They will not put their reputation in danger by adding any xyz search technologies.
    2. Good technologies must be backed by a strong (and innovative) business model.
    3. If any one of these models can reach market threshold Google will be happy to buy them or crush them in weeks (may be!).

    
    I might be inexperienced compared to other R/W writers, I can see following two ways of beating google.

    1. MSN & Yahoo search don't use great search technologies, how come they are still Number 2 & 3? These search engines were promoted by giants like MS and Y!. Even minnows like AOL search comes within top 10 (right?). I really see a possibility of disturbing the search market dynamics if people focused giants like Apple or IBM or NASA can promote above mentioned serach engines. Reasons - these companies have more reputation than google, can influence mass, deep pockets, could be a social contribution (...)

    2. Well am working on it :)

    Happy New Year!

    Dhruba

    Posted by: Dhruba Baishya | January 7, 2007 4:09 AM





15. Not sure "best" next-generation search can be segmented so cleanly. A number of new engines employ a combination of techniques. Take Kosmix (http://www.kosmix.com) which significantly outperforms Google for discovery oriented searches. You can see more details via my link to Lightspeed's blog (full disclosure - I'm an investor). Kosmix has developed algorithmic/AI techniques to impose a taxonomy on they way they index the web but combine this with "social search" thru UGC and editorial-based category tags which power the navigation engine. The net effect is a very compelling experience for online research/discovery versus the linear result stream produced by Google.

    Posted by: Ravi Mhatre | January 7, 2007 7:46 PM





16. Google is now worth more than 50 BILLION dollars. Love 'em or hate 'em, it's going to be a long, long time before any new technology, GUI or algorithm will be able to go up against Research and Development and Advertising/Marketing budgets that are backed by billions...

    Posted by: HeSaid. | January 12, 2007 3:18 PM