The Thirst for Knowledge…

Out-Googling Google.

by David J. Bilinsky

July 29, 2008.

♫ I tried my best to let you know
That I’m not trying to test you
It’s just so hard to let you go
When I have nothing against you…♫

Words and music by Midtown

It was just a matter of time. It isn’t that I have anything against you, Google, indeed I have learned so much from you, but in time, all things must end. Alas, I have found another.

Cuil (pronounced cool) is the new well, Cool search engine in town. And she is big – very big – so big that the developers of Cuil claim that she searches three times as many web pages as Google and ten times as many as Microsoft. When you are looking for a reference, it is reassuring that your search engine goes deep – very deep.

But it isn’t just the depth of the search that is determinative. Cuil claims that it ranks pages not based on popularity of a web page, but rather on their content and relevance. This has important consequences. Since popularity is not on their radar screen, there is no need to retain personal search histories. Your search history is always private!

The way information (Cuil is the irish word for knowledge) is displayed is also different, too. Rather than seeing Google Adwords on the right, you see a box “Explore by Category” in the upper right hand side of the screen. This box contains a listing of subjects related to your search and you can roll-over them to see related information displayed relevant to your search. If you ‘hover’ over an ‘Explore by Category’ item, a definition will also appear.

Searches are displayed in an array; each cell containing information on the ‘hit’. The searches typically contain graphics as well as information on the hit as well as the required web link. Furthermore, Cuil displays “Tabs” of information, each of which contains other choices closely related to your search. When you submit a query, Cuil displays a list of suggestions allowing you to look before you leap.

On the Cuil website, there is a handy “Add Cuil to Firefox” link to place Cuil inside the quick search box on the toolbar (which for me includes Wikipedia, Merriam-Webster and other ‘must have’ links).

So – Google – old buddy, old pal – it isn’t that I have anything against you..

Comments

  1. Well I hear you, but the site is just too fast to be that deep – and lo and behold when I ran one search and was told 14,000 odd hits, I was surprised when it ran out less than 150 hits in.

    On the good side, it did include blog postings.

    As for an alternative try Copernic, and then in the preferences set the default for the depth of search from 10 to 250 hits per engine, and tick the maximum number of engines. Copernic slows to a crawl, but the results when they come are far deeper and are stacked by relevance with duplicates excluded.

    What I also like is that one can then run searches within the search set.

    Nice rhetoric Cuil, though the Guardian is sceptical – and why a name drawn from nook or cranny in Gaelic?

  2. Alejandro Manevich

    Simon,

    Do you find Copernic preferable to Google for certain types of tasks? If so, is this because of the depth of its search or because of its functionality?

    I haven’t used Copernic in many years, but I was under the (perhaps mistaken) impression that Google had long won the battle. Metasearch engines like Copernic and Dogpile seemed to be trendy in the 1990s, when search engines were in their infancy, but do they now offer anything that a single, well-designed search engine does not?

    Alex

  3. I find it interesting that you mention Copernic. It has been around for some time, if my memory serves me correctly I believe that it predates google (correct me if I’m wrong). But…. Copernic seems to run counter to the trend of online services, in that it is a piece of search software that one downloads. It seems that people keep searching for a google replacement, what is the Slaw opinion on the ability of a Copernic to make a comeback based on depth of results and downloadable software?

  4. The point is that one uses it as a supplement to Google (which is my everyday engine, though I do use Live.com from time to time).

    Copernic comes into its own when one needs to be more confident that you’ve looked under more stones. As I tell younger lawyers when speaking about search engines, no engine is more than partial – see searchenginewatch.com generally – and different engines will raise unique results which are invisible even from Google.

    I use Copernic when I want to be more definitive – and can wait to pull an expresso, while the search is being run.

    For those of us who grew up with five minute waits for results from old Lexis and QL, that’s acceptable. I can’t imagine anyone under thirty however, waiting more than five seconds for a screen to load or a result to materialize.

  5. These results are not cuil, terrible in fact!

    Try a simple search for our PM Stephen Harper. #1 is a wikipedia article on a fictional character of the same name. Really… it should be able to have our pm.gc.ca page come up number one.

    Second example, tried a search of my own name. Cuil inserted photos OVER the results – none of which were me! The two people licking each other, well, I won’t dignify…

    I don’t use copernic, but have had it installed in the past with good results. My preference after Google is actually Yahoo. There are many times I could argue that Yahoo results are better than Google, but for the fact of limited search-share, I would be tempted to switch.

    Cuil? big waste of GC funding so far. And 5-10 years too late.

  6. Tried a few test searches, and was not impressed overall. For example, a search for “Federal Court of Appeal Canada” gave rather disappointing results on Cuil. Google’s first link is the FCA homepage. I think the algorithms will need some work before I switch my primary search engine…

  7. Figuring out SEO for Google has been a fine-tuned art that takes a long time to figure out.

    A newer search engine using different algorithms is not something I’m particularly happy about. I quick look around and it definitely produces more results. It also seems blog-heavy, which is good for us involved in the space. But there’s also a lot of redundancy (and bugs), and I already see areas that could be easily exploited.

    Finally, I take some comfort in knowing that no matter how much better it may (or not) be than Google, consumer habits mean that it’s not going to change any time soon, irrespective of whether the lat Per Bak was right about self-organized criticality (1996).

  8. Well, first off Cuil is well, different. One would expect this if it is to be a real alternative to Google or the rest of the other search paradigms.

    Yes it does display info in an array format. I rather like that, but it is more a personal preference related to design.

    The search algorithms may need some further refinement, but after all it does search differently from Google and I think Google has focused our way of thinking to date.

    But perhaps the strongest point is the lack of any search history in Cuil. As lawyers I think we should be concerned about on-line privacy and should support efforts to deliver useful information without having to trade-off our bits of our personal privacy in return. After all, if you go into a library and do research, there is no record of the information that you have referenced if you just browse. If you use Copernic, for example, with its hundreds of sources – are you leaving electronic fingerprints all over the web?

    I say give Cuil a chance to adapt and respond. This is just the initial release. I would expect that initiatives like Cuil will spawn other innovations that will lead to more meaningful search methods outside of the shadow of Big Brother.

  9. So guys, are you trying to scare away our new contributor on the first day or what?!

    This discussion is very interesting. I agree there could be potential to Cuil. And the privacy issue is a very important one. Cuil was started by former Google employees I believe, so they must be very aware this is an issue. Unfortunately they got off to a very bumpy start, and as a result all the fabulous word of mouth traffic they generated could work against them as people came, saw, and decided they weren’t interested.

    Having blogs so thoroughly indexed is a mixed blessing. The regular Google search currently serves up only two hits from any one site, so you get a real mix of sites but not an idea of how much applicable content is available from that site. On the other hand with Cuil, they do not seem to place that restriction on the search, so you could get any number of hits from one site. In the case of a blog with hundreds of posts, you may have to search through all the blog posts to find other content (search my name and you will see what I mean–good luck slogging through those 900+ blog posts from the past 4 years!).

    I am interested to see what they do with this start. How will they overcome the negative impression they generated?

    In the meantime, you’ve done a great job kicking off your presence here, David.

    Cheers,
    Connie

  10. Alejandro Manevich

    Just one minor addendum to Connie’s post, for those who may not be aware: one can remove the 2 hit per site limitation on Google searches (and do many other wonderful things) by using a javascript bookmarklet.

    Alex

  11. If Cuil is relying on ‘content’ and not hyperlinks, it may be aiming to be the search of choice in the semantic web. Google relies on the careful attention web writers pay to the hyperlinks they create. I think the idea of the semantic web is to generate the same standardization and reliability when it comes to the content of a page.

  12. Alejandro: Thanks! I like that there are different bookmarklets depending upon your preference for number of hits. Too bad we have to resort to an outside app for this though–the average person definitely won’t be using bookmarklets.

    Michael: very interesting thought.

  13. The search on my name using Cuil was not inpressive. It found my blog (homepage) 4 times in the first 6 pages of results, as well as my ClaimID page twice (about the amount of times I’ve visited it). As mentioned the pictures it inserts would be nice, were it not for the fact they usually have no connection to what they are next to.
    The “Explore by Category” in the upper right hand side of the screen – a nice idea, but as it has already identified that there are 2 scott vine’s who are librarians and one who is a snowboarder, I’m not sure why I’d want to explore native American Writers, as it suggests.

    I wish them luck, and some people are saying it works quite well for academic searching, but I think they should have left this locked up until it worked – not have it in everlasting beta (like Goggle) but also not launch making claims it can not match.

  14. …and can i just hang my head in shame for my unforgivable ‘their’ as opposed to there …

  15. No need if you have a friend with edit rights. All fixed Scott. :)