Open-Source Search Engine
I learned about Nutch.org from a comment on Douwe Osinga’s post entitled “Why We Should Leave Google Behind.” They’re making a stab at creating an open-source, Google-quality search engine.
It’s an ambitious undertaking, but Objects Search is already putting the Nutch engine to work — together with the open source Carrot2 Project — on searches for the web, news (including blogs and RSS feeds), images, and more. The results it returns look promising, though the image search is pleagued by broken images.
Objects Search is definitely something to keep an eye on. One of its best qualities is that you can click on an “explain” link to get details on how and why a site was ranked the way it was. Since the engine is open and provides info on how results were ranked, search engine optimization would become much easier than it currently is with Google’s opaque PageRank algorithm. On the other hand, gaming the system might be extraordinarily easy.
For a more literal example of an open source search engine, check out Koders. It searches the source code of a bunch of open source software projects.
Update: CNet, Tech News World, and Wikipedia have more to say about Nutch. There are some big names in the technology world involved, such as Tim O’Reily.
Yahoo! Research Labs has its own implementation of Nutch running, but the results suck in comparison to Objects Search. It turns up more pages, but the relevance is awful. The paranoid among us might suspect that Yahoo was attempting to discredit their potential open-source competitor…
Line and paragraph breaks are automatic. Your e-mail address will never be published publicly unless you put it in your comment (and then I’d probably edit it out).
Please read my comment policy if this is your first time commenting here.