The score used to rank a page is computed as the document rank score + a relevance score. This score is then combined with a proximity score if the query has more than two terms using reciprocal rank fusion. Below are some bonuses which may be applied to the document rank and relevance scores.

Host Keyword Bonus

Potential bonus to add to relevance score. The number of occurrences of search term divided by the number of host name keywords is the fraction of this bonus that will be added to the relevance score.

Title Bonus

Potential bonus to add to relevance score. The number of occurrences of search term divided by the number of words in the title is the fraction of this bonus that will be added to the relevance score.

Path Bonus

Potential bonus to add to relevance score. The number of occurrences of search term divided by the number of words in the path portion of the url is the fraction of this bonus that will be added to the relevance score.

Proximity Bonus

Proximity scores for multi-term queries are normalized between 0 and 1, this is the weighting factor to multiply that basic score before adding it to the overall score used for ranking

CLD Url Bonus

Bonus to add to doc rank score if the url is a company level domain.

Host Url Bonus

Bonus to add to doc rank score if the url is a a hostname.

User Rank Bonus

User rank scores (created by making a classifier) for a document are normalized between 0 and 1, this is the weighting factor to multiply that basic score before adding it to the overall score used for ranking.

Wiki Bonus

Bonus to add to doc rank score if the url points to a wikipedia page.

Num Slashes Bonus

Bonus to add to doc rank score based on the number of '/' present in the url, based on the idea that the lesser the count, the closer the url is to the root page.