New Search Engine Cuil Tries to Take on Google
"Cuil (pronounced COOL) is a search engine that combines the largest Web index with content-based relevance methods, organized results, and complete user privacy. The company's next generation approach to search is the result of proprietary breakthroughs in search architecture and ranking algorithms."
Ciul claims to have an index of 120 billion pages, which it states is "three times more than any other search engine." We have no reason to doubt this claim, but in our limited hands-on with Cuil this morning, Cuil often either turned up fewer results than Google or no results at all. (For instance, separate searches for the words "cpu," "hamlet," and "quagmire" turned up no results.) As the site is new, the empty results could be a function of first-day bugs or user overload. (Although, curiously, while searching for "French" turned up no results, "French bread" produced 250 results. Searching for "French bread" on Google produced 1,520,000 results.)
Unlike Google's PageRank approach to ranking pages by perceived importance (or "popularity" as Cuil calls it), Cuil claims to rank "results by the content on each page," while examining the context of the search term on the pages:
"Cuil searches the Web for pages with your keywords and then we analyze the rest of the text on those pages. This tells us that the same word has several different meanings in different contexts. Are you looking for jaguar the cat, the car or the operating system?"
For instance, when we searched for "French bread," the results page included tabs with more context-specific examples of our search, such as "French Bread Recipe" and "Sourdough French Bread." Also, an "Explore by Category" panel was on the results page with subject-related topics such as "New Orleans Cuisine" and "American Sandwiches."
| Click on the image above to see the full-sized image.|
Cuil was founded by husband and wife Tom Costello and Anna Patterson:
"Mr. Costello researched and developed search engines at Stanford University and IBM; Ms. Patterson is best known for her work at Google, where she was the architect of the company’s large search index and led a Web page ranking team."