If you've done a search on thestar.com in the past few weeks, you may have noticed an improvement in the quality and depth of our search results.
The biggest change is that we are now indexing new articles as soon as we publish them to the site -- in other words, breaking news is now searchable, so you always be able to access the latest story updates through our search engine.
But the improvements don't stop there.
You can sort your results by date (oldest first or newest first) or by relevancy based on the relative frequency in the articles of the keyword you searched.
In addition to finding articles published within the past seven days, our new search results include topic pages, such as Sports or Business or special reports such as our Afghanistan page.
You'll also find more information in our search results, including bylines, the original published source of the material -- the print edition of the Toronto Star or thestar.com -- and links to more articles from our columnists.
If you don't find what you're looking for in our 7-day search, one click will carry over your search terms to our paid archive service with is free to search. And a new search page -- search.thestar.com -- gives you other options for searching Star photos and pages.
Most importantly, our search engine handles text searching much better, using boolean logic to locate stories that include one or more of the keywords you enter.
The software behind our improved searching is a Java-based engine called Lucene (read a long analysis of Lucene here), developed by Doug Cutting a senior architect at Excite.com. Lucene is another example of the flexibility -- and lower cost -- that open source software has brought to web development and commercial websites such as the Star.
Our searches still aren't perfect. We're working on improving the relevancy rankings to make them more meaningful and to offer more sources of content.
Future phases will also focus on provided targetting searching of movies, car reviews and recipes.
Looks Like U're Nice Search Engine is Broken this morning...
HTTP Status 404 - /SearchWeb/Search.do
type Status report
message /SearchWeb/Search.do
description The requested resource (/SearchWeb/Search.do) is not available.
Apache Tomcat/5.5
Posted by: retired 905 area code elected Halton Councillor Robert Heaton | October 27, 2006 at 06:02 AM