Hacker News new | ask | show | jobs
by Typhon 2545 days ago
Here's an idea : do the 0.1 version by crawling and indexing the web in a perfectly standard way, apply the usual, run-of-the-mill search algorithms, then diff it against the Bing/Google/etc first results which will be biased towards commercial stuff. The higher a page is among their results, the higher the probability that it's commercial in some sense.
1 comments

So...you want to build a Google except each query starts on the last page of results?
No, it's just a simple heuristic to get started.