July 28, 2008
New Search Engine - Cuil
Being a person who has been tracking search engines and writing about them for many years, my ears went BOING! and perked right up. Sounds really interesting!
Cuil (pronounced "cool" or "kewl") is supposed to have the largest search engine database ever. So far, so good.
They address one of the really big recent concerns with Google, too - privacy.
I know certain people this is really going to attract! Very important, and glad to see them do this.
And the interface? OMG, made me salivate, I swear. This is GOOD stuff! A lot of other search engines could learn important lessons from Cuil when it comes to the interface. They have beautifully integrated a number of recent trends in interface design - clustering, semantic search, visual thumbnails.... They even have two layers of clusters - the tabbed topics across the top and the suggested terms for narrowing a topic over at the right. Here is a screenshot of a sample search on the word "diabetes".
I liked the way they selected the most important topics for the right hand topical clusters and gave the option to expand to include more. Mousing over a topic gives more detail, so you can easily browse in by topic to a really decent level of granularity.
I just loved that when you mouse over the bottom level suggested search terms Cuil even provides a definition! w00t!
Where this all broke down was in the implementation. I tried clicking through the DIET section and selecting the subtopics. Of the four suggested -- low carbohydrate diet, low fat diet, dieting, and fasting -- only two of them were actually valid for search. I'd show you a screenshot, but their servers are down just this sec. ("We’ll be back soon... Due to overwhelming interest, our Cuil servers are running a bit hot right now. The search engine is momentarily unavailable as we add more capacity. Thanks for your patience.") I was curious because two of the four topics suggested are actually bad ideas for diabetics, at least the last I heard ("low-fat diet" and "fasting"). Indeed, those were the two suggested topics that, when clicked upon, tried to add the search terms to the search, and came back with zero results. I was baffled why the search engine would suggest inclusion of nonviable concepts.
The next concern that popped up for me was when I clicked through one of the subtopics that did work.
Personally, I rather enjoy spiritual reading and websites, however, I cannot consider them a top level source of content on most healthcare topics. I confess I was more than a bit concerned that for a search on "diabetes low-carbohydrate diet" Cuil returned as the top result a web page on Alzheimers from the BibleLife organization. Uh, oops?
I have already gone on record as having concerns about Google losing their edge for search result relevance as SEO has become more effective, so I wondered if the results would be similar in other search engines. I popped over to Google and tried running the same search.
Ahhhhh, much better. :) Leading research journals and professional organizations on diabetes seem much more relevant to my mind. So it isn't that better results weren't available, it is just that for some inexplicable reason, Cuil felt BibleLife's pages were more appropriate than the American Diabetes Association professional publications. I think some tweaking is in order here. I went back and dug a little more, and suspect that this is related to what they say on their main Information page*: "Rather than rely on superficial popularity metrics, Cuil searches for and ranks pages based on their content and relevance." Perhaps they should consider some way to include authority or credibility in that analysis?
*BTW, Cuil? The link to "About Us" on the main page is broken. Was tricky to find this.
So what do I really think? Here is the short and sweet from the discussion at Plurk.
Posted by pfa at July 28, 2008 08:50 AM