Following structure summarizes output of AnchorSpamPenalizer. Spammy anchors are tagged with LINK_SPAM_PHRASE_PENALIZER and demoted to SPAM locality in anchor-localizer.cc

Attributes

• anchorEnd (type: integer(), default: nil) - End date of the last anchor of the document.
• anchorFraq (type: number(), default: nil) - Ratio of spam demoted period to all anchor period.
• anchorStart (type: integer(), default: nil) - Start date of the first anchor of the document.
• demoted (type: integer(), default: nil) - Following field record details of anchor demotion in action. How many anchors were demoted.
• demotedAll (type: boolean(), default: nil) - Demoted all anchors in the period or only anchors classified as spam.
• demotedEnd (type: integer(), default: nil) - End date of the demotion period.
• demotedStart (type: integer(), default: nil) - Start date of the demotion period.
• phraseCount (type: number(), default: nil) - Following fields record signals used in anchor spam classification. How many spam phrases found in the anchors among unique domains.
• phraseDays (type: number(), default: nil) - Over how many days 80% of these phrases were discovered.
• phraseFraq (type: number(), default: nil) - Spam phrases fraction of all anchors of the document.
• phraseRate (type: number(), default: nil) - Average daily rate of spam anchor discovery.
• processed (type: integer(), default: nil) - Total number of processed anchors.
• sampled (type: boolean(), default: nil) - True if anchors were sampled during observation phrase.
• sources (type: list(GoogleApi.ContentWarehouse.V1.Model.IndexingDocjoinerAnchorTrustedInfo.t), default: nil) - Detailed information about trusted sources and match computation. Populated only when --anchorspam_penalizer_debug=true.
• spamDebugInfo (type: String.t, default: nil) - Additional debug information about computation of spam probability.
• spamPenalty (type: number(), default: nil) - Combined penalty for anchor demotion.
• spamProbability (type: number(), default: nil) - Predicted probability of spam.
• trustedDemoted (type: integer(), default: nil) - Number of trusted anchors used in computation of spam probability.
• trustedExamples (type: String.t, default: nil) - Examples of trusted sources.
• trustedMatching (type: integer(), default: nil) - Number of trusted anchors with anchor text matching spam terms.
• trustedTarget (type: boolean(), default: nil) - Following fields record details about trusted anchors True if is this URL is on trusted source.
• trustedTotal (type: integer(), default: nil) - Total number of trusted sources for this URL.