I had an idea for an alternative search engine a few years ago thats a bit simpler to implement than this one. First we extract all the external links from Wikipedia dumps. Then we ingest only those sites into the index. The entire database comprises sites that have already been screened by Wikipedians. Gaming Wikipedia is generally more difficult than gaming Google or Bing. In theory, SERPs would be of a less commercial and more substantive nature than those we get from Google and Bing.