TrustRank

TrustRank is a link analysis algorithm used by search engines like Google to combat web spam and improve search result quality. It works by manually identifying a set of high-quality, trustworthy "seed" pages. Human evaluators determine these seed sites, ensuring they are reputable and not involved in spamming activities.

The algorithm then calculates a "trust score" for other pages on the web by measuring their link distance from these seed pages. The core principle is that good pages tend to link to other good pages, and spam pages tend to link to spam pages (or vice versa, but spam pages are less likely to receive links from genuinely reputable sites).

Therefore, pages linked to directly from seed pages receive a high trust score. This trust score then propagates through the web along outbound links. However, with each link distance from the seed set, the trust score diminishes. Pages that are many links away from a trusted seed page receive a lower trust score.

This score helps search engines differentiate between legitimate websites and spam sites. Pages with high TrustRank are more likely to be ranked higher in search results, while pages with low TrustRank may be demoted or even removed from the index.

TrustRank is often used in conjunction with other anti-spam algorithms and ranking factors. It is considered a labor-intensive process due to the manual selection of seed pages, but it is effective at identifying and penalizing link farms and other spam techniques designed to artificially inflate a website's ranking.