From 8fdff2748ffbc4ce3d36c75f2bd33182b641e895 Mon Sep 17 00:00:00 2001 From: Eugen Rochko Date: Fri, 16 Aug 2019 01:24:03 +0200 Subject: Add more accurate account search (#11537) * Add more accurate account search When ElasticSearch is available, a more accurate search is implemented: - Using edge n-gram index for acct and display name - Using asciifolding and cjk width normalization on display names - Using Gaussian decay on account activity for additional scoring (recency) - Using followers/friends ratio for additional scoring (spamminess) - Using followers number for additional scoring (size) The exact match precedence only takes effect when the input conforms to the username format and the username part of it is complete, i.e. when the user started typing the domain part. * Support single-letter usernames * Fix tests * Fix not picking up account updates * Add weights and normalization for scores, skip zero terms queries * Use local counts for accounts index, adjust search parameters * Fix mistakes * Using updated_at of accounts is inadequate for remote accounts --- app/chewy/accounts_index.rb | 36 +++++++ app/models/account.rb | 6 ++ app/models/account_stat.rb | 2 + app/services/account_search_service.rb | 189 ++++++++++++++++++++++----------- 4 files changed, 169 insertions(+), 64 deletions(-) create mode 100644 app/chewy/accounts_index.rb (limited to 'app') diff --git a/app/chewy/accounts_index.rb b/app/chewy/accounts_index.rb new file mode 100644 index 000000000..e11b80039 --- /dev/null +++ b/app/chewy/accounts_index.rb @@ -0,0 +1,36 @@ +# frozen_string_literal: true + +class AccountsIndex < Chewy::Index + settings index: { refresh_interval: '5m' }, analysis: { + analyzer: { + content: { + tokenizer: 'whitespace', + filter: %w(lowercase asciifolding cjk_width), + }, + + edge_ngram: { + tokenizer: 'edge_ngram', + filter: %w(lowercase asciifolding cjk_width), + }, + }, + + tokenizer: { + edge_ngram: { + type: 'edge_ngram', + min_gram: 1, + max_gram: 15, + }, + }, + } + + define_type ::Account.searchable.includes(:account_stat), delete_if: ->(account) { account.destroyed? || !account.searchable? } do + root date_detection: false do + field :id, type: 'long' + field :display_name, type: 'text', analyzer: 'edge_ngram', search_analyzer: 'content' + field :acct, type: 'text', analyzer: 'edge_ngram', search_analyzer: 'content', value: ->(account) { [account.username, account.domain].compact.join('@') } + field :following_count, type: 'long', value: ->(account) { account.active_relationships.count } + field :followers_count, type: 'long', value: ->(account) { account.passive_relationships.count } + field :last_status_at, type: 'date', value: ->(account) { account.last_status_at || account.created_at } + end + end +end diff --git a/app/models/account.rb b/app/models/account.rb index 60c06aaf0..392cc625f 100644 --- a/app/models/account.rb +++ b/app/models/account.rb @@ -127,6 +127,8 @@ class Account < ApplicationRecord delegate :chosen_languages, to: :user, prefix: false, allow_nil: true + update_index('accounts#account', :self) if Chewy.enabled? + def local? domain.nil? end @@ -169,6 +171,10 @@ class Account < ApplicationRecord subscription_expires_at.present? end + def searchable? + !(suspended? || moved?) + end + def possibly_stale? last_webfingered_at.nil? || last_webfingered_at <= 1.day.ago end diff --git a/app/models/account_stat.rb b/app/models/account_stat.rb index 9813aa84f..6d1097cec 100644 --- a/app/models/account_stat.rb +++ b/app/models/account_stat.rb @@ -16,6 +16,8 @@ class AccountStat < ApplicationRecord belongs_to :account, inverse_of: :account_stat + update_index('accounts#account', :account) if Chewy.enabled? + def increment_count!(key) update(attributes_for_increment(key)) end diff --git a/app/services/account_search_service.rb b/app/services/account_search_service.rb index e1874d045..2d602a31d 100644 --- a/app/services/account_search_service.rb +++ b/app/services/account_search_service.rb @@ -4,105 +4,134 @@ class AccountSearchService < BaseService attr_reader :query, :limit, :offset, :options, :account def call(query, account = nil, options = {}) - @query = query.strip - @limit = options[:limit].to_i - @offset = options[:offset].to_i - @options = options - @account = account + @acct_hint = query.start_with?('@') + @query = query.strip.gsub(/\A@/, '') + @limit = options[:limit].to_i + @offset = options[:offset].to_i + @options = options + @account = account - search_service_results + search_service_results.compact.uniq end private def search_service_results - return [] if query_blank_or_hashtag? || limit < 1 + return [] if query.blank? || limit < 1 - if resolving_non_matching_remote_account? - [ResolveAccountService.new.call("#{query_username}@#{query_domain}")].compact - else - search_results_and_exact_match.compact.uniq - end + [exact_match] + search_results end - def resolving_non_matching_remote_account? - offset.zero? && options[:resolve] && !exact_match? && !domain_is_local? - end + def exact_match + return unless offset.zero? && username_complete? - def search_results_and_exact_match - return search_results.to_a unless offset.zero? + return @exact_match if defined?(@exact_match) - results = [exact_match] + @exact_match = begin + if options[:resolve] + ResolveAccountService.new.call(query) + elsif domain_is_local? + Account.find_local(query_username) + else + Account.find_remote(query_username, query_domain) + end + end + end - return results if exact_match? && limit == 1 + def search_results + return [] if limit_for_non_exact_results.zero? - results + search_results.to_a + @search_results ||= begin + if Chewy.enabled? + from_elasticsearch + else + from_database + end + end end - def query_blank_or_hashtag? - query.blank? || query.start_with?('#') + def from_database + if account + advanced_search_results + else + simple_search_results + end end - def split_query_string - @split_query_string ||= query.gsub(/\A@/, '').split('@') + def advanced_search_results + Account.advanced_search_for(terms_for_query, account, limit_for_non_exact_results, options[:following], offset) end - def query_username - @query_username ||= split_query_string.first || '' + def simple_search_results + Account.search_for(terms_for_query, limit_for_non_exact_results, offset) end - def query_domain - @query_domain ||= query_without_split? ? nil : split_query_string.last - end + def from_elasticsearch + must_clauses = [{ multi_match: { query: terms_for_query, fields: likely_acct? ? %w(acct) : %w(acct^2 display_name), type: 'best_fields' } }] + should_clauses = [] - def query_without_split? - split_query_string.size == 1 - end + if account + return [] if options[:following] && following_ids.empty? - def domain_is_local? - @domain_is_local ||= TagManager.instance.local_domain?(query_domain) - end + if options[:following] + must_clauses << { terms: { id: following_ids } } + elsif following_ids.any? + should_clauses << { terms: { id: following_ids, boost: 100 } } + end + end - def search_from - options[:following] && account ? account.following : Account - end + query = { bool: { must: must_clauses, should: should_clauses } } + functions = [reputation_score_function, followers_score_function, time_distance_function] - def exact_match? - exact_match.present? - end + records = AccountsIndex.query(function_score: { query: query, functions: functions, boost_mode: 'multiply', score_mode: 'avg' }) + .limit(limit_for_non_exact_results) + .offset(offset) + .objects + .compact - def exact_match - return @exact_match if defined?(@exact_match) + ActiveRecord::Associations::Preloader.new.preload(records, :account_stat) - @exact_match = begin - if domain_is_local? - search_from.without_suspended.find_local(query_username) - else - search_from.without_suspended.find_remote(query_username, query_domain) - end - end + records end - def search_results - @search_results ||= begin - if account - advanced_search_results - else - simple_search_results - end - end + def reputation_score_function + { + script_score: { + script: { + source: "(doc['followers_count'].value + 0.0) / (doc['followers_count'].value + doc['following_count'].value + 1)", + }, + }, + } end - def advanced_search_results - Account.advanced_search_for(terms_for_query, account, limit_for_non_exact_results, options[:following], offset) + def followers_score_function + { + field_value_factor: { + field: 'followers_count', + modifier: 'log2p', + missing: 1, + }, + } end - def simple_search_results - Account.search_for(terms_for_query, limit_for_non_exact_results, offset) + def time_distance_function + { + gauss: { + last_status_at: { + scale: '30d', + offset: '30d', + decay: 0.3, + }, + }, + } + end + + def following_ids + @following_ids ||= account.active_relationships.pluck(:target_account_id) end def limit_for_non_exact_results - if offset.zero? && exact_match? + if exact_match? limit - 1 else limit @@ -113,7 +142,39 @@ class AccountSearchService < BaseService if domain_is_local? query_username else - "#{query_username} #{query_domain}" + query end end + + def split_query_string + @split_query_string ||= query.split('@') + end + + def query_username + @query_username ||= split_query_string.first || '' + end + + def query_domain + @query_domain ||= query_without_split? ? nil : split_query_string.last + end + + def query_without_split? + split_query_string.size == 1 + end + + def domain_is_local? + @domain_is_local ||= TagManager.instance.local_domain?(query_domain) + end + + def exact_match? + exact_match.present? + end + + def username_complete? + query.include?('@') && "@#{query}" =~ Account::MENTION_RE + end + + def likely_acct? + @acct_hint || username_complete? + end end -- cgit