Comments on: The Greatest Debate: Algorithms vs. Interfaces http://altsearchengines.com/2007/09/21/the-greatest-debate-algorithms-vs-interfaces/ The most wonderful search engines you've never seen! Sat, 30 Aug 2008 00:37:04 +0000 http://wordpress.org/?v=2.2 By: NitinK http://altsearchengines.com/2007/09/21/the-greatest-debate-algorithms-vs-interfaces/#comment-10485 NitinK Wed, 26 Sep 2007 13:48:41 +0000 http://altsearchengines.com/2007/09/21/the-greatest-debate-algorithms-vs-interfaces/#comment-10485 @Babak: You're absolutely right - HTML is really focused on text and <i>display markup</i>, so it's a very inefficient way for search engines to gather information from the web. Gathering structured data like RSS (or even from semantic processing of regular web pages, if it can be done) will be far more powerful! It would allow regular search users to write parameterized, database-like queries. I must disagree with you, however, on the importance of the <b>size</b> of the crawled index. I can give you counter-examples of search engines which work well with far smaller indexes than that of Google. In the end, the user only really cares about finding the few, *relevant* answers to her query (and I don't believe those are directly correlated). Counter-examples include: Mahalo, which explicitly addresses only the top 10K search queries, yet has gotten significant traction AND could easily be considered a superlative search engine for those common queries. Similarly, vertical search engines, by definition, have dramatically smaller indexes, yet can be both popular and highly effective. @Babak: You’re absolutely right - HTML is really focused on text and display markup, so it’s a very inefficient way for search engines to gather information from the web. Gathering structured data like RSS (or even from semantic processing of regular web pages, if it can be done) will be far more powerful! It would allow regular search users to write parameterized, database-like queries.

I must disagree with you, however, on the importance of the size of the crawled index. I can give you counter-examples of search engines which work well with far smaller indexes than that of Google. In the end, the user only really cares about finding the few, *relevant* answers to her query (and I don’t believe those are directly correlated).

Counter-examples include: Mahalo, which explicitly addresses only the top 10K search queries, yet has gotten significant traction AND could easily be considered a superlative search engine for those common queries. Similarly, vertical search engines, by definition, have dramatically smaller indexes, yet can be both popular and highly effective.

]]> By: Charles Knight http://altsearchengines.com/2007/09/21/the-greatest-debate-algorithms-vs-interfaces/#comment-10432 Charles Knight Wed, 26 Sep 2007 00:27:37 +0000 http://altsearchengines.com/2007/09/21/the-greatest-debate-algorithms-vs-interfaces/#comment-10432 Babak, Thank you for that comment. You're right, we must be willing to look in all directions. Of the "Top 100" Alternative Search Engines, who do you think is on the right track? Charles Babak,

Thank you for that comment. You’re right, we must be willing to look in all directions.

Of the “Top 100″ Alternative Search Engines, who do you think is on the right track?

Charles

]]> By: Babak http://altsearchengines.com/2007/09/21/the-greatest-debate-algorithms-vs-interfaces/#comment-10426 Babak Tue, 25 Sep 2007 22:40:45 +0000 http://altsearchengines.com/2007/09/21/the-greatest-debate-algorithms-vs-interfaces/#comment-10426 It is not a new interface or a new algorithm that can effectively compete with Google; it is where you search and how you crawl. HTML code was not invented so it can be crawled effectively by search engines; it is just so primitive and inefficient. Structured XML data such as RSS feeds is where we can start to effectively compete with Google. Google is as new to this area as other start-ups. Still new start-ups have to collaborate and innovate constantly to stay ahead of curve. Even Google admits that it is not able to efficiently crawl and index this huge amount of data that is constantly changing on Internet. Structured data is fundamentally scalable, efficient, self-explanatory and platform independent. It offers the basis for better search algorithms and AI applications and its adaptation to custom built interfaces is just extraordinary. BH It is not a new interface or a new algorithm that can effectively compete with Google; it is where you search and how you crawl. HTML code was not invented so it can be crawled effectively by search engines; it is just so primitive and inefficient. Structured XML data such as RSS feeds is where we can start to effectively compete with Google. Google is as new to this area as other start-ups. Still new start-ups have to collaborate and innovate constantly to stay ahead of curve.

Even Google admits that it is not able to efficiently crawl and index this huge amount of data that is constantly changing on Internet. Structured data is fundamentally scalable, efficient, self-explanatory and platform independent. It offers the basis for better search algorithms and AI applications and its adaptation to custom built interfaces is just extraordinary. BH

]]> By: iAdvert.mobi » Weekly Wrapup, 17-21 September 2007 http://altsearchengines.com/2007/09/21/the-greatest-debate-algorithms-vs-interfaces/#comment-10232 iAdvert.mobi » Weekly Wrapup, 17-21 September 2007 Mon, 24 Sep 2007 06:31:39 +0000 http://altsearchengines.com/2007/09/21/the-greatest-debate-algorithms-vs-interfaces/#comment-10232 [...] did a debate this week on Algorithms vs. Interfaces, a hot topic in search blogs right now. ASE editor Charles Knight says that Google has the most [...] […] did a debate this week on Algorithms vs. Interfaces, a hot topic in search blogs right now. ASE editor Charles Knight says that Google has the most […]

]]> By: Charles Knight http://altsearchengines.com/2007/09/21/the-greatest-debate-algorithms-vs-interfaces/#comment-10197 Charles Knight Sun, 23 Sep 2007 18:51:59 +0000 http://altsearchengines.com/2007/09/21/the-greatest-debate-algorithms-vs-interfaces/#comment-10197 NK, Truthfully, I think Google could replicate almost anything they want to replicate. They're into space exploration for Gosh sake! The $65,000 question is, what can be done about it? If all of the best Vertical Alts combine to take money away from Google, Google can't pick them off or re-create dozens of the top verticals. This has always been my thought: "Collaborate or Perish." *Together* they would be too much to copy. And then add a new Algo, and they're in a fix. I do feel that they have a "blind spot" which is exactly why Ford, GM & Chrysler are now such miserable failures. CK NK,

Truthfully, I think Google could replicate almost anything they want to
replicate.

They’re into space exploration for Gosh sake! The $65,000 question is,
what can be done about it?

If all of the best Vertical Alts combine to take money away from Google,
Google can’t pick them off or re-create dozens of the top verticals.

This has always been my thought: “Collaborate or Perish.” *Together*
they would be too much to copy. And then add a new Algo, and they’re
in a fix.

I do feel that they have a “blind spot” which is exactly why Ford, GM &
Chrysler are now such miserable failures.

CK

]]> By: GadgetGadget.info - Gadgets on the web » The Greatest Debate: Algorithms vs. Interfaces http://altsearchengines.com/2007/09/21/the-greatest-debate-algorithms-vs-interfaces/#comment-10116 GadgetGadget.info - Gadgets on the web » The Greatest Debate: Algorithms vs. Interfaces Sun, 23 Sep 2007 07:55:50 +0000 http://altsearchengines.com/2007/09/21/the-greatest-debate-algorithms-vs-interfaces/#comment-10116 [...] unknown wrote an interesting post today!.Here’s a quick excerptOne week ago, Nitin Karandikar wrote this post defending the pursuit of a new algorithm to beat Google. It’s not too long, so I encourage you to please stop, go back, and read it before you go any further. … [...] […] unknown wrote an interesting post today!.Here’s a quick excerptOne week ago, Nitin Karandikar wrote this post defending the pursuit of a new algorithm to beat Google. It’s not too long, so I encourage you to please stop, go back, and read it before you go any further. … […]

]]> By: Weekly Wrapup, 17-21 September 2007 : Forecast-Blog http://altsearchengines.com/2007/09/21/the-greatest-debate-algorithms-vs-interfaces/#comment-10114 Weekly Wrapup, 17-21 September 2007 : Forecast-Blog Sun, 23 Sep 2007 06:18:45 +0000 http://altsearchengines.com/2007/09/21/the-greatest-debate-algorithms-vs-interfaces/#comment-10114 [...] did a debate this week on Algorithms vs. Interfaces, a hot topic in search blogs right now. ASE editor Charles Knight says that Google has the most [...] […] did a debate this week on Algorithms vs. Interfaces, a hot topic in search blogs right now. ASE editor Charles Knight says that Google has the most […]

]]> By: NitinK http://altsearchengines.com/2007/09/21/the-greatest-debate-algorithms-vs-interfaces/#comment-9944 NitinK Sat, 22 Sep 2007 03:25:14 +0000 http://altsearchengines.com/2007/09/21/the-greatest-debate-algorithms-vs-interfaces/#comment-9944 Very well written, Charles! You've got me convinced - well, almost! :-) You make some great points. The dazzling, attractive new user interfaces that the Alts are creating are indeed unique, creative and valuable, and likely to draw attention. But, to use your own argument, if one or more of these user interfaces really takes off and starts drawing users away, how long would it take the Google hordes to reproduce it? A Google team could retrofit a cool new UI on top (possibly under a different domain) just as easily as a third-party can, once they know what they're shooting for. If you want an example, I give you <a href="http://www.searchmash.com" rel="nofollow">SearchMash</a> - a Google property. On the other hand, changing the underlying search algorithm - say based on Semantic Search or NLP - is a great deal harder, IMHO. However, I must admit that I'm biased - being a software developer at heart, and "art-challenged" to boot! [Until my wife started buying clothes for me, my idea of variety was various shades of gray ...] I can't wait to see what Kaila has to say! Very well written, Charles! You’ve got me convinced - well, almost! :-)

You make some great points. The dazzling, attractive new user interfaces that the Alts are creating are indeed unique, creative and valuable, and likely to draw attention. But, to use your own argument, if one or more of these user interfaces really takes off and starts drawing users away, how long would it take the Google hordes to reproduce it? A Google team could retrofit a cool new UI on top (possibly under a different domain) just as easily as a third-party can, once they know what they’re shooting for. If you want an example, I give you SearchMash - a Google property.

On the other hand, changing the underlying search algorithm - say based on Semantic Search or NLP - is a great deal harder, IMHO.

However, I must admit that I’m biased - being a software developer at heart, and “art-challenged” to boot! [Until my wife started buying clothes for me, my idea of variety was various shades of gray …]

I can’t wait to see what Kaila has to say!

]]>