Tag: TG_UNUSED

Page Rank. The factor will be remarked.

The complex Static Rank is assembled from static components according to a separate formula ((http://wiki.yandex-team.ru/jandekspoisk/kachestvopoiska/obshayAfermula/tekushhiekomponenty/#oftnd1 *))).

The value of the news detector calculated in the Hippo. Always 0 with a detector value less than the threshold.

The value of the cosinematchmaxprediction factor for the Stryim ALICEMUSIC

The number of incoming links. Remembrance.

The value of the factor annotationmaxvalueEWEEGTED for Stryim ALICEMUSIC

The relative frequency of the words in the links (1 - the words of the request are often found in links, 0.3 - rarely); More precisely, the value of this factor is pessimized provided: TR = 0 && LR = 0 & (there is not a single link with all the words of the request) && (did not pass the quorum) && (at least one pair of words of the request is found in the text)

The percentage of incoming links with the words of the request

The percentage of incoming links with all the words of the request

The value of the commerce detector calculated in the Hippo.

The value of the Factor AnnotationMatchweighhedValue for Stryim ALICEMUSIC

Weight of maximum coincidence of forms in the text and request

Remapped mascot feature TrafgraphOutAllSE_share_m

Remapped mascot feature NoExtClicksShare

Search traffic - transitions from search engines to the site (2nd formula)

Search traffic - transitions from search engines to the site (2nd formula)

BCLMIXPLAINK000001 Factor value for Alicemusic

Three levels of coincidence of the region of links and request

The value of the CMMATCHTCHTOP5AVGMATCH factor for the Stryim ALICEMUSIC

PERWORDCMMAXMATCHMIN Factor value for Alicemusic Stryim

The value of the factor attenv1_bm15_k05 for the Stryim ALICEMUSIC

The value of the factor annotationmaxvalueEWEEGTED for Stryim ALICEMUSIC

The weight of the words of the request that is in the Links

The value of the factor attenv1_bm15_k05 for the Stryim ALICEMUSIC

ALLWCMMAXMatch factor

The value of the cosinematchmaxprediction factor for the Stryim ALICEMUSIC

Matrixnet is applied to all factors - formula (tg_unized - to prevent the entrance to any formulas)

The maximum number of forms in all words of the request

Summer of the number of forms balanced by scales

Undested amount of the number of forms

The value of the CMMATCHTCHTOP5AVGMATCH factor for the Stryim ALICEMUSIC

Quality of incoming links (hauser classifier) ​​corrected

Quality classifier of incoming links 2 corrected

The value of the CMMATCHTCHTOP5AVGMATCHVALUE factor for Stryim ALICEMUSIC

The value of the PERWORDCMMAXPREDITICIONMIN factor for the Stryim ALICEMUSIC

Ukrainian Page Rank

The frequency of links to the site

The number of almost-periodic links

DSSM model with early binding, trained on reforming and learned by ASR hypotheses of musical requests for Alice

DSSM model with early binding, trained on reforming and learned on musical requests for Alice

The value of the pirate detector calculated in the hippo.

The host of the document is recognized in the request

URL consists only of the host, which is recognized in the request

URL is the plot of Yandex News

URL feature computed from rapid clicks spy_log counters with decay of 1 day

URL feature computed from rapid clicks spy_log counters with decay of 0.5 days

They are considered as (80 - x) / 80, where X is the age of the document in the watch. Factors make sense only for the fast -button base (the last 80 hours). Not used in ranking. Used in disconnecting.

They are considered as (80 - x) / 80, where X is the age of the document in the watch. Factors make sense only for the fast -button base (the last 80 hours). Not used in ranking. Used in disconnecting.

Medical host quality for new marks.

Medical host quality for new marks for experiments.

Finance or law host quality for new marks.

Finance or law host quality for new marks for experiments.

Finance or law host quality for new marks.

Finance or law host quality for new marks for experiments.

Factor for host in list of documentation cs hosts for experiments

URL feature computed from rapid clicks spy_log counters with decay of 3 days

BM25 with different parameters for different fields, including an incoming anchortekst. The weight of the text of the links included on the page is normalized depending on Delta Page Rank links

URL feature computed from rapid clicks spy_log counters with decay of 14 days

The service factor that was needed to search the site, and in the future it will still be needed.

Age of rapid clicks spy_log update, in seconds

Freshness of rapid clicks spy_log update

The value of the chopped commerce detector, calculated in the hippo.

Probabilistic model built on the texts of incoming links

'Country praets' (AUXQC)

'Country praets' (AUXQC)

Logarithm of the number of shingles in the document added by the owner of the site as original texts in ((http://wiki.yandex-team.ru/jandekspoisk/jekosistema/marketingPr/webmasters/plan/vtorcontect of originality plugin)). It does not participate in the formula, it is needed to disconnect the takes

The average filtered number of sources of authorship of the document. It does not participate in the formula, it is needed to disconnect the takes

Document links dispersion

The amount of factors QueryDownerClicksFRC and BM25FDPRFIXED with scales 0.358449 and 0.184922, respectively. '565' in the name of the factor does not need to be perceived literally, it is Legashi or a typo.

The amount of factors 192 and 341 with scales 0.298942 and 0.454625, respectively.

Dispersion of languages ​​in XMAP

It is known about URL that it is shown too often with very low relevance (according to Bert and/or BM25)

The ratio of the number of incoming links, the text of which is the URL, is one of the incoming links

The length of the URL with an accuracy to the symbol. Disconnected in production.

BM25 page title by texts of links to it

The number of COO-Thrilling links between hosts

The value of the health detector calculated in the Hippo.

OffersBase feature for ecoboost.

OffersBase feature for ecoboost.

OffersBase feature for ecoboost.

OffersBase feature for ecoboost.

The share of unique trigrams in the trigrams of links

Share of unique trigrams of links in trigrams header

URL document corresponds to the city of the user

The sum of the maximum values ​​of Sourcerank's for each incoming link, taking into account the uniqueness of the owner.

Allows you to evaluate whether the document is 'alive' is from the point of view of links to it coming.

The core of the page of the pages on which there is a metric counter

OffersBase feature for ecoboost.

OffersBase feature for ecoboost.

Business kernel.

Business kernel.

Business kernel.

URL feature computed at the request time from rapid clicks search counters with decay of 1 day

URL feature computed at the request time from rapid clicks search counters with decay of 30 days

Personalized Turkish Pagerank

URL is an offer in the latest version of the market base.

User outflow coefficient from the search after a visit to the site

The document contains a name from the request.

Url satisfies the regexp expression set in the demon

Num of Sovetnik URLS

The factor is used in Selectionrank. TG_UNUSED: should not be included in the formulas to avoid feedback

URL feature computed at the request time from rapid clicks search counters with decay of 3 days

URL feature computed from rapid clicks search frozen counters with decay of 1 day

URL feature computed from rapid clicks search frozen counters with decay of 30 days

URL feature computed from rapid clicks search frozen counters with decay of 3 days

Aries is a service

Host feature computed at the request time from rapid clicks spy_log counters with decay of 3 days

Host feature computed at the request time from rapid clicks spy_log counters with decay of 3 days

Host feature computed at the request time from rapid clicks spy_log counters with decay of 14 days

Host feature computed at the request time from rapid clicks spy_log counters with decay of 14 days

Host feature computed at the request time from rapid clicks spy_log counters with decay of 3 days

Host feature computed at the request time from rapid clicks spy_log counters with decay of 14 days

Host feature computed from rapid clicks spy_log counters with decay of 3 days

Host feature computed from rapid clicks spy_log counters with decay of 3 days

Host feature computed from rapid clicks spy_log counters with decay of 14 days

Host feature computed from rapid clicks spy_log counters with decay of 14 days

Host feature computed from rapid clicks spy_log counters with decay of 3 days

Host feature computed from rapid clicks spy_log counters with decay of 14 days

Is legal video distributor

Average value of feature OneProductProbability

Average value of feature ManyProductsProbability

Average value of feature PayDetectorPredict

Aries is a partner

The document is Shopinshop

The value of the conversion of the request calculated in the Hippo.

For Ural from Ytier, it is known that he has low -quality content

For Ural from Ytier, it is known that he has a content of acceptable quality

For Ural from Ytier, it is known that he has good quality content

For Urla from Ytier, it is known that he has an excellent content content

The host has an ecom purchase.

The host has a purchase by Visit Log.

URL is a product on the market.

URL is a product on the market and has Offerid.

URL is Shopinshopcpa.

Owner has an ecom purchase.

Owner has a purchase by Visit Log.

DSSM Prediction of the probability of URL + Title that the document is an overlap.

In the offer from the new Parser, the PartnerOfferContent Available field is exhibited.

In an offer from the new Parser, the PartnerOfferContent Available Field == True.

The value of PURCHASETOTALPredict, calculated in the Hippo.

The value of serpsummarysurpluspredict, calculated in the hippo.

Requestwith120d3ClickpartPredict value, calculated in the Hippo.

The value of the requester detector of the parasites calculated in the hippo.

FIN_LAW.DSSM Predictions URL + TITLE.

Predict Cruelty.dssm URL + TITLE models.

The value of Halfecompredict, calculated in the Hippo.

The value of Matrixnet is a slow ranking model.

The value of Matrixnet with a quick ranking model.

The value of the Matrixnet filter model.

The value of Polynom is a slow ranking model.

The value of Polynom is a quick ranking model.

The value of the Polynom filter ranking model.

A sign that the document was received by machine translation

Prediction of Med_with_Trash.DSSM (Medic. Document model with Tresh Valley in Lern) Models for URL + Title.

Prediction FIN_LAW_WITH_TRASH.DSSM (Fin-Jur. Document model with a tresh valve in Lern) Models for URL + Title.

The value of Matrixnet with a fast filter model.

The value of Polynom with a quick filter ranking model.

Meet Matrixnet Meaning.

Meeting Polynom.

A document is a short video (Tiktok, Reels, Shorts).

Document-telegram channel in web format.

Document - post in a telegram.

DSSM CLOSE DISCOUNT according to data specific for Alice

The prediction of the total time spent to the end of the session, provided that this pair is implemented by the request-document

The prediction of the contribution of this pair request-document to the timetable

The prediction of the percentage of the length of the track, which will be lost subject to the implementation of this pair of the request

Neural model of content quality for medical subjects

Neural model of content quality for medical subjects (for ex -)

Neural model of content quality for financial and legal topics

Neural model of content quality for financial and legal topics (for exposures)

Random float in [0,1] by user request and document

Neural model of content quality for SOS topics

Neural model of content quality for SOS subjects (for ex -)

Prediction of the time of the session, provided that this pair is requested by the request-document

Document - this is a selection of Tiktok /Tag

Document - this is a selection of Tictock /Discovery

Document - this is a selection of Tiktok /Music

URL is a channel/fasting from a verified social network account

DSSM, predicting whether the site is a facial

Document DSSM model Language Classifier Rus.

Document DSSM model Language Classifier Eng.

Document DSSM model Language Classifier Other.

DSSM Prediction to determine Alice's irrelevant answers

DSSM prediction (URL + Title), trained for the Page_QUALYY signal and implemented in RTHUB, the first slot.

Type of canonized Urla Yandex Music - track

Document has Fio from original request

PAGE Quality 1 experiments factor

Page Quality 2 experiments factor

Rating of news agency from agencies.json (Yandex.News resource)

Page Quality 3 experiments factor

Page Quality 4 experiments factor

Page Quality 5 experiments factor

The document came from Webtier1

The document came from Platinum0

Neural document model for finding unexpected tin

Medical host quality fresh.

DSSM prediction of google specificity for query

1 if host include js from google-analytics.com

1 if host include js from googleapis.com

Host speed estimation

Medical host quality for metric.

The document came from Quickmed

Medical host quality.

The document has a turbo page for Desktop platforms. Updates on top of the base are delivered through SaAS.

Page quality aggregated by host (avg).

The request was recognized as having intent to copyright objects protected by anti-Pirate memorandum.

Transferred Ethos predicted classifier for marking on the relevance of video.

Reduced ETHOS predictive classifier, trained in the synthetic sample 'Request characteristic of the pirate site' VS 'Request is characteristic of the site far from this'

In this slot, there was no zero feature

DSSM regression for the separation of memorandum and no requests

Transferred Ethos predicted classifier trained to distinguish a memorandum queries from a random

DSSM regression of embeds to separate Pirato specific and no requests

The document came from Quick but not from QuickRt

The document came from QuickRt

The document came from Callisto

The quality indicator of the site from the point of view of factors about user behavior aggregated to the owners.

Bans of Antispam from erf

Bans on gsm of Antispam from erf

Bans on fresh of Antispam from erf

The document has a turbo page for Mobile platform. Updates on top of the base are delivered through SaAS.

The document has a turbo page for Desktop platforms. Updates on top of the base are delivered through SaAS.

'Random' factor for commercial sites.

Neuron document model for finding unexpected tin (for ex -)

The value of the freshness detector calculated in the hippo. Always 0 with a detector value less than the threshold.

Max tokens frequency in music scenarios

Max tokens frequency in video scenarios

Average tokens frequency in music scenarios

Average tokens frequency in video scenarios

Max tokens frequency in music scenarios (without high frequency tokens)

Max tokens frequency in video scenarios (without high frequency tokens)

Average tokens frequency in music scenarios (without high frequency tokens)

Average tokens frequency in video scenarios (without high frequency tokens)

Max tokens frequency in video select from gallery scenario

Avg tokens frequency in video select from gallery scenario

Max tokens frequency in general conversation scenarios

Avg tokens frequency in general conversation scenarios

Max tokens frequency in search scenarios

Avg tokens frequency in search scenarios

Max tokens frequency in other scenarios

Avg tokens frequency in other scenarios

Max tokens bigrams frequency in music scenarios

Max tokens bigrams frequency in video scenarios

Average tokens bigrams frequency in music scenarios

Average tokens bigrams frequency in video scenarios

Max tokens bigrams frequency in music scenarios (without high frequency tokens)

Max tokens bigrams frequency in video scenarios (without high frequency tokens)

Average tokens bigrams frequency in music scenarios (without high frequency tokens)

Average tokens bigrams frequency in video scenarios (without high frequency tokens)

Max tokens bigrams frequency in video select from gallery scenario

Avg tokens bigrams frequency in video select from gallery scenario

Max tokens bigrams frequency in general conversation scenarios

Avg tokens bigrams frequency in general conversation scenarios

Max tokens bigrams frequency in search scenarios

Avg tokens bigrams frequency in search scenarios

Max tokens bigrams frequency in other scenarios

Avg tokens bigrams frequency in other scenarios

The value of the freshness detector calculated in the hippo. Always 0 with a detector value less than the threshold.

The value of the news detector calculated in the Hippo. Always 0 with a detector value less than the threshold.

The value of the pirate detector calculated in the hippo.

The value of the commerce detector calculated in the Hippo.

The value of the chopped commerce detector calculated in the hippo.

The value of the health detector calculated in the Hippo.

The value of Purchasetotalpredict, calculated in the Hippo.

The value of serpsummarysurpluspredict, calculated in the hippo.

Requestwith120d3ClickpartPredict value, calculated in the Hippo.

The value of Halfecompredict, calculated in the Hippo.

The value of the conversion of the request calculated in the Hippo.

The value of the requester detector of the parasites calculated in the hippo.

Classifier of one product in the request

Classifier of several goods in the request

The value of the Proxima DCG-5, calculated in the hippo.

The request is fresh

DSSM prediction of google specificity for query

The request was recognized as having intent to copyright objects protected by anti-Pirate memorandum.

Transferred Ethos predicted classifier for marking on the relevance of video.

Reduced ETHOS predictive classifier, trained in a synthetic sample 'Request characteristic of the pirate site' VS ', the request is characteristic of the site far from this'

In this slot, there was no zero feature

DSSM regression for the separation of memorandum and no requests

Transferred Ethos predicted classifier trained to distinguish a memorandum queries from a random

DSSM regression of embeds to separate Pirato specific and no requests

Manual formula of the classifier of the pa*of the request of the request

Manual formula of the classifier M*Morandum of request

Time remaining before a sports event between several teams

The difference between the current date, and the date in the request (1 - a request for today's date, 0 - about the date after 10 days and older)

DSSM Prediction of money spent on a request

The prediction of the classifier is that the request is white (that is, it is not worth showing tin on it)

DSSM prediction of the complexity of the request

Indicator - is the request of the SEO -request

Indicator - is the request of the SEO -request from the list No. 1

Indicator - is the request of the SEO request from the list No. 2

Request contains full name from FIO Rules

The prediction of the classifier is that upon request you need to show URL from a mobile market

Prediction of the classifier to especially commercial requests

Prediction of the CP Classifier Requests

Classifier of a request for equipment repair services: office, household, computers, phones

Whether the request is porn.

Spare factor for the rapid adding functionality

Spare factor for the rapid adding functionality

Spare factor for the rapid adding functionality

Spare factor for the rapid adding functionality

Spare factor for the rapid adding functionality

Spare factor for the rapid adding functionality

Prediction of the CP Classifier Requests

Prediction of the requesting classifier of violence

Prediction of the requesting classifier of weapons or drugs

Musical request for the script of Arab Alice

Classifier of a request for pornographic for Indonesian

Classifier of requests for the requirement only English -speaking pages

Classifier of requests for the requirement of mixed English -speaking and Russian -speaking pages

The classifier of requests for the requirement of either only English -speaking pages, or English -speaking and Russian -speaking pages together

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Dummy feature

Activity (shows+clicks+diplics) of the user in OID. Suspended

P+K+user DC in ORG1 Requests to TOP-1. Suspended

P+K+user's DC in the Request section. Suspended

P+K+DC user in the section of the document. Suspended

P+K+user DC in the network in the request. Suspended

P+K+DC user to the network of a document. Suspended

User activity in Byak. Weighed as well

User activity in the male. Weighed as well

User activity in Navi. Weighed as well

xFactor collaborative filtering on OID

xFactor collaborative section

xFactor collaborative filtering on the network

P+K+DC user in the section of the document. Suspended for the popularity of the heading and prescription

Organization in the status of Unreliable

Organization in the status of ClosedTemporary

Organization not in the status unlILLIBLE, CLOSEDPERMENNENTLY, CLOSEDTEMPORARY

[0-1] distance from DOC to Ull /500 km

Function of the exchange of rating at a distance.

Flag boost Lee Category

The presence of priority placement

Document position when sorted by L1 relevance. Only calculated for learning pools.

Restrictions related to quarantine

The mode of work with a closed hall for visitors

The values ​​of the click formula

Quorum values ​​for hits in attributes of signs

How much we want to get this advertiser to fulfill our promises on advertising

Kvorum values ​​for hits in attributes of signs for the vertical Request Window

Kvorum values ​​for hits in attributes of signs for the vertical What Were

In this factor, information about pessimization is put in

A random additive to a faster is placed in this factor

Pessimize, unwanted organization

In this factor, information about Busts is placed

Beating by the presence of tripadvisor urla

Bewilder Tripadvisor Urla in the entertainment section

Bay by rating in Tripadvisor

Buster by the number of reviews in Tripadvisor

Price rating for sorting by price (more expensive)

Spring number of the best object

The length of the best object

The proximity of the request to the DSSM BERT -based lawn cluster

The proximity of the request to the DSSM BERT -based vacuum cleaners cluster

The ratio of orders to shows based on a request as NID and a document as Categid

CTR based on a request as QueryDopp and a document as a Hyperid with a click function for Threshold

The opposite to the CTR based on the request as MarketqueryDopphidnid and the document as Shopid with a click function for Threshold

CTR based on a request as QueryDopp and a document as a mskuid with a click function for Threshold

CTR based on a request as iCookie and a document as a Hyperid with a click function for Threshold

Rational sigmoid number of clicks with a click function on the Threshold 120 based on a request as MarketqueryDopphidnid and a document as Categid

The ratio of clicks with the CPA click function by the ThreShold to clicks with a click function by a request based on a request as a NID and a document as Categid

The ratio of orders to clicks with a click function for DwellTime based on a request as MarketqueryDopPhidnid and a document as a Hyperid

CTR based on a request as MarketQuerydopphidnid and a document as a Categid with a click function for Threshold

The ratio of clicks with the click function by Threshold 120 to orders based on a request as NID and a document as Categid

CTR based on a request as MarketQuerydopphidnid and a document as a Hyperid with a click function for Threshold

The ratio of clicks with the CPA click function for Threshold to orders based on a request as MarketqueryDopphidnid and a document as Categid

The ratio of clicks with a click function by DwellTime to clicks with CPA CLICHING FOR THE WHECHILD based on a request as MarketqueryDopPhidnid and a document as Categid

CTR based on a request as MarketQuerydopphidnid and a document as a Hyperid with a click function for Threshold 120

CTR based on a request as a NID and a document as a Categid with a click function for Threshold

Rational sigmoid number of clicks with the CPA CLICE ON THE THE WHOSED BURGENT BUILD AS THE MARKETQURYDOPPHIDNID and Document as CATEGID

Rational sigmoid number of clicks with the CPA CLICE ON THE THE WHECHOOLD based on a request as MarketqueryDopPhidnid and Document as Hyperid

CTR based on a request as iCookie and a document as a Categid with a click function for Threshold

Rational sigmoid number of clicks with a CPA CLIC CLOSE CPADWELTIME based on a request as MarketqueryDopPhidnid and a document as a Hyperid

Rational sigmoid of the number of clicks with a click function by a request based on a request as MarketqueryDopphidnid and a document as Categid

Rational sigmoid number of clicks with a click function based on a request as a MarketqueryDopphidnid and document as Categid

The ratio of clicks with the CPA click function by CPadwellTime to clicks with the CPA click function based on the request as NID and the document as Categid

The ratio of orders to clicks with the CPA click function based on a request as a MarketqueryDopphidnid and a document as Categid

Rational sigmoid number of orders based on a request as MarketqueryDopphidnid and a document as Categid

Rational sigmoid number of shows based on a request as MarketqueryDopphidnid and a document as a Hyperid

Rational sigmoid number of shows based on a request as MarketqueryDopphidnid and a document as Categid

The ratio of clicks with the click function by Threshold 120 to clicks with the CPA function for the request based on the request as MarketqueryDopPhidnid and the document as Categid

Rational sigmoid number of shows based on a request as Querydopp and a document as WareMD5

The attitude of orders to clicks with a click function for Threshold based on a request as MarketqueryDopphidnid and a document as Categid

Rational sigmoid number of shows based on a request as HID and a document as Hyperid

0 in rtmodels_factors.txt

1 in rtmodels_factors.txt

2 in rtmodels_factors.txt

3 in rtmodels_factors.txt

4 in rtmodels_factors.txt

5 in rtmodels_factors.txt

6 in rtmodels_factors.txt

7 in rtmodels_factors.txt

8 in rtmodels_factors.txt

9 in rtmodels_factors.txt

10 in rtmodels_factors.txt

11 in rtmodels_factors.txt

12 in rtmodels_factors.txt

13 in rtmodels_factors.txt

14 in rtmodels_factors.txt

15 in rtmodels_factors.txt

16 in rtmodels_factors.txt

17 in rtmodels_factors.txt

18 in rtmodels_factors.txt

19 in rtmodels_factors.txt

L1 top position

DSSM prediction of google specificity for query

Images porn classifier experimental formula result

The document contains url from assessments (not to use in training)

Images porn classifier experimental formula result

The document contains url from assessments (not to use in training)

The length of the URL with an accuracy to the symbol. Disconnected in production.

Logarithm of the number of shingles in the document added by the owner of the site as original texts in ((http://wiki.yandex-team.ru/jandekspoisk/jekosistema/marketingPr/webmasters/plan/vtorcontect of originality plugin)). It does not participate in the formula, it is needed to disconnect the takes

The average filtered number of sources of authorship of the document. It does not participate in the formula, it is needed to disconnect the takes

Search traffic - transitions from search engines to the site (2nd formula)

Search traffic - transitions from search engines to the site (2nd formula)

1 if host include js from google-analytics.com

1 if host include js from googleapis.com

The number of COO-Thrilling links between hosts

L2 factor similar to Doppurlosctrlog5Decay30 only with fixed OS on iOS

All factors on L3 are applied by the matrixet formula.

All factors on L2 are applied by the matrixet formula.

Last position where the document was the worst in a pair.

Doc is quite good to be boosted

The value of the polynomial formula of the thast.

Season of a serial object.

Serial object series.

The language of the document strictly corresponds to the domain (fine English)

The average value of the Additionrank factor is average.

The average value of the Additionrank factor in PRS

The average deviation of the Additionrank factor in PRS

CoEF-T of Pieron Variation for Additionrank factor in PRS

The proximity of the request and document on the Neuronet model trained on DSSM and static/request factors.

If a word from a request is found in Ann adds a normalized weight of the word to this factor 1 time.

Probability of the REL label from the toloka

Probability of the NOT_REL label from the toloka

Probability of the 404 label from the toloka

Is online stream

Date of issue of the document (for TV shows).

Predict formula for the relevance of a fresh document.

In this factor (into this index), it seems to be rewritten in an unknown place, presumably from some dense Legacy Code

The position of the document in L1 top. Used to define targett in teaching a dynamic L1 Top model

Trailer from KinoPoisk

The muzzle of the document

Non -accuracy of the document

ReleaseDateRaw (unused in formula)

Skor model of duplicate prediction.

Score models of relevance.

Was the document viewed by the user on the filter pn_vhs#ft_search_seen

Was the document viewed by the user on the filter pn_vhs#ft_spy_seen

Was the document viewed by the user on the filter pn_vhs#ft_tvonline_player_seen

It costs 1 when the request for several tags and the document has already got into the tag

The number of fire reactions of the user on TEGU

The number of haha ​​reaction user on TEGU

The number of wow-reactions of the user on TEGU

Number of SAD-reactions of the user on TEGU

The number of Angry reaction user on TEGU

The number of Fire-reactions of the user on TEGU with half-life 7 days

The number of haha ​​reaction of the user on TEG with half-life 7 days

The number of wow-reactions of the user on TEGU with half-life 7 days

The number of the user's SAD-reactions for TEGU with half-life 7 days

The number of Angry-reactions of the user on TEG with half-life 7 days

The number of Fire-reactions of the user on TEG with half-life 30 days

The number of haha ​​reaction of the user on TEG with half-life 30 days

The number of wow-reactions of the user on TEGU with half-life 30 days

The number of the user SAD-reactions for TEGU with half-life 30 days

The number of Angry reaction user on TEG with half-life 30 days

The number of Fire-reactions of the user on TEG with half-life 2 hours

The number of haha ​​reaction of the user on TEG with half-life 2 hours

The number of wow-reactions of the user on TEGU with half-life 2 hours

The number of the user's SAD-reactions for TEGU with half-life 2 hours

The number of Angry reaction user on TEG with half-life 2 hours

The number of Fire-reactions of the user on TEG with half-life 8 hours

The number of haha ​​reaction of the user on TEG with half-life 8 hours

The number of wow-reactions of the user on TEGU with half-life 8 hours

The number of the user SAD-reactions for TEGU with half-life 8 hours

The number of Angry reaction user on TEG with half-life 8 hours

Number of Heart-reactions of the user on TEGU

The number of bored-reaction user on TEGU

The number of the user of the TEGU user with half-life 7 days

The number of bored-reaction user for TEG with half-life 7 days

The number of the user of the TEGU user with half-life 30 days

The number of bored-reaction user for TEG with half-life 30 days

The number of the user's Heart reaction of TEG with half-life 2 hours

The number of bored-reaction user for TEG with half-life 2 hours

The number of the user's Heart reaction of TEG with half-life 8 hours

The number of bored-reaction user for TEG with half-life 8 hours

The document was added as a candidate from the dock2dok to a viewed document

The share of the user on the Search logs of this tag

The share of the user on the Search logs of this tag

The share of the user's LVT Search logs of this tag in 30 days

The share of the user on the Search logs of this tag in 30 days

The share of the user on Ether Logs of this tag

The share of the user on Ether logs of this tag

The share of the user's lvt by Ether logs of this tag in 30 days

The share of the user on Ether logs of this tag in 30 days

The share of the user on the spy logs of this tag

The share of the user on the SPY logs of this tag

The share of user LVT for SPY logs of this tag in 30 days

The share of the user on the SPY logs of this tag in 30 days

The share of emoji angry pressed by user

The share of emoji Dislike pressed for a tag

The share of emoji fire pressed for a tag

The proportion of emoji haha ​​closed by user for tag

The share of emoji like pressed for a tag

The share of emoji sad for a tag closed by user

The share of emoji wow for a tag closed by user

The number of Tag subscribers on Yutyub

The number of Tag Views on Yutyub

The number of video tags on yutub

KL-condition between the crypto distribution of the user by income and the corresponding suspended distribution of TEG over thirty-day ethereal TVT

KL-condition between the crypto-distribution of the user by age and the corresponding balanced distribution of TAG along the thirty-day ethereal TVT

KL-condition between the crypto-distribution of the user on the floor and the corresponding suspended distribution of TAG along the thirty-day broadcast TVT

DOT Product Embeddding Tag and Laikh user

DOT Product Embedding Tag and Dislayer User

Dot Product between (pn_entity#et_ ranks, v0 embeditions on bikes on the user's air and tag

Dot Product between (pn_entity#et_ ranks, v1 embeditions on bikes on the user's air and tag

The category is dedicated to the video block

The category is dedicated to films

The category is dedicated to the series

The category is dedicated to sports

DOT Product pn_vhs#et_history_dssm user embedding and category

The share of emoji heart for a tag closed by user

The share of Emoji Skip for a tag closed by user

The share of emoji bored -pressed user for tag

The number of documents with this tag

DOT Product between CF# user embeditions and category

Cos(Agg(DocProfile, TotalViewTimeProfileWeightExtractor(Search)), UserProfile, PN_VHS#ET_USER_2_DOC, RT_Sum)

Whether the user pressed the cross on the categories

Whether the user is signed for the category

Cos(Agg(DocProfile, TotalViewTimeProfileWeightExtractor(Ether)), UserProfile, PN_VHS#ET_USER_2_DOC, RT_Sum)

Cos(Agg(DocProfile, TotalViewTimeProfileWeightExtractor(Spy)), UserProfile, PN_VHS#ET_USER_2_DOC, RT_Sum)

Cos(Agg(DocProfile, LogViewTimeProfileWeightExtractor(Search)), UserProfile, PN_VHS#ET_USER_2_DOC, RT_Sum)

Cos(Agg(DocProfile, LogViewTimeProfileWeightExtractor(Ether)), UserProfile, PN_VHS#ET_USER_2_DOC, RT_Sum)

Cos(Agg(DocProfile, LogViewTimeProfileWeightExtractor(Spy)), UserProfile, PN_VHS#ET_USER_2_DOC, RT_Sum)

Cos(Agg(DocProfile, ReleaseTimestampProfileWeightExtractor(RT_Sum_90d)), UserProfile, PN_VHS#ET_USER_2_DOC, RT_Sum)

Cos(Agg(DocProfile, ReleaseTimestampProfileWeightExtractor(RT_Sum_30d)), UserProfile, PN_VHS#ET_USER_2_DOC, RT_Sum)

Cos(Agg(DocProfile, ReleaseTimestampProfileWeightExtractor(RT_Sum_7d)), UserProfile, PN_VHS#ET_USER_2_DOC, RT_Sum)

Logarithm of the number of new documents in 1 day

Logarithm of the number of new documents for 3 days

Logarithm of the number of new documents for 7 days

Logarithm of the number of new documents for 30 days

DOT Product pn_vhs#et_history_dssm user embedding and category V1

The popularity of category based on TVT

The popularity of category based on TVT

The popularity of category based on TVT

The popularity of category based on TVT

The popularity of category based on TVT

DOT Product Embedding Tag and Skipa user

DOT Product Embeddings Tag and User by Clicks

DSSM v0 categ embedding

DSSM v1 categ embedding

vhs_history_outer

The amount of Ether TVT documents divided by the number of documents

Ether TVT amount for 30 days of documents divided by the number of documents

Ether TVT amount for 7 days of documents divided by the number of documents

Ether TVT amount for 8 hours of documents divided by the number of documents

Search TVT amount divided by the number of documents

Search TVT amount for 30 days of documents divided by the number of documents

Search TVT amount for 7 days of documents divided by the number of documents

Search TVT amount for 8 hours of documents divided by the number of documents

SPY TVT amount divided by the number of documents

SPY TVT amount for 30 days of documents divided by the number of documents

SPY TVT amount for 7 days of documents divided by the number of documents

SPY TVT amount for 8 hours of documents divided by the number of documents

Aggregation for categories et_crypta_doc

The number of authors in this category

The share of muzzle documents

The presence of an avatar in the public

A number of unique words in a request.

Sigmoid, the sum of the deviations of the terms of the request

Sigmoid, the sum of the number of news articles for all query terms

Sigmoid, the sum of the 'peaks' of the query terms

The maximum (normalized) deviation for the term request, refined by sigmoid

The maximum number of news articles for 72 hours for the term of request for the term of request

The maximum 'peak' maximum maximum

The share of unique words from the request through the Terms, to which the response from the RTMR table was received

Search traffic - transitions from search engines to the site (2nd formula)

Search traffic - transitions from search engines to the site (2nd formula)

Search traffic - transitions from search engines to the site (2nd formula)

Search traffic - transitions from search engines to the site (2nd formula)

The factor is used in Selectionrank. TG_UNUSED: should not be included in the formulas to avoid feedback

The factor is used in Selectionrank. TG_UNUSED: should not be included in the formulas to avoid feedback

Logarithm of the number of shingles in the document added by the owner of the site as original texts in ((http://wiki.yandex-team.ru/jandekspoisk/jekosistema/marketingPr/webmasters/plan/vtorcontect of originality plugin)). It does not participate in the formula, it is needed to disconnect the takes

The average filtered number of sources of authorship of the document. It does not participate in the formula, it is needed to disconnect the takes

The length of the URL 'and with an accuracy to the symbol. Disconnected in production.

The core of the page of the pages on which there is a metric counter

Logarithm of the number of shingles in the document added by the owner of the site as original texts in ((http://wiki.yandex-team.ru/jandekspoisk/jekosistema/marketingPr/webmasters/plan/vtorcontect of originality plugin)). It does not participate in the formula, it is needed to disconnect the takes

The average filtered number of sources of authorship of the document. It does not participate in the formula, it is needed to disconnect the takes

The length of the URL 'and with an accuracy to the symbol. Disconnected in production.

The core of the page of the pages on which there is a metric counter

The number of COO-Thrilling links between hosts

The number of COO-Thrilling links between hosts

Rating of news agency from agencies.json (Yandex.News resource)

Rating of news agency from agencies.json (Yandex.News resource)

Random by (userId, timestampMinutes / 5) feature

Position in the top l1

In the case of the presence of concrete, 1 is affixed upon their passage, otherwise 0.5, in the absence of 0

Resid metafactor on SR(24)

set feature to 1 when url owner in &rearr/scheme_Local/SetBoostOwners/Owners=‘host1.ru,host2.ru’

A random feature is considered to be hash (Ownerid, Timestamp). It is calculated only for the flow ru.commercial (Note: TG_UNUSED is used to gain in the default pools)

'Random' factor for commercial sites.

Thematic business cakel

Is host a mobile store.

Is host a mobile store for query platform.

From reseved_owner namespace in SaaS-KV

From reseved_url namespace in SaaS-KV

From reseved_owner namespace in SaaS-KV for fast reaction

From reseved_url namespace in SaaS-KV for fast reaction

bert model which predicts 1rel 2vital 0other for arabic alice music search scenario

slot for support of two parallel aramusic models

small-bert-model sbr:1721013142 predict Predict_click_pers_ce_1534051635_standartized_mse_target Distillation of the click pers ce bert model 1534051635

small-bert-model sbr:1721013142 predict Predict_click_odd_mse_1512795366_standartized_mse_target Distillation of the click odd mse bert model 1512795366

split-bert-model sbr:2005741938 predict Predict_xLarge_prediction_nov_lav_1990027087_standartized_mse Split distillation of the xlarge model with relev10 1990027087

split-bert-model sbr:2005741938 predict Predict_Large_predict_target_fin_law_1967416198_standartized_mse Split distillation of the large model 1967416198

split-bert-model sbr:2005741938 predict Predict_Large_predict_target_med_doc_1967182618_standartized_mse Split distillation of the large model 1967182618

split-bert-model sbr:2005741938 predict Predict_Large_predict_target_sos_1969262884_standartized_mse Split distillation of the large model 1969262884

bert model 2021/09/boyalex/2405042847 head: Predict_jul_xlarge_sinsig_baskets_2339818891_stand_mse

bert model 2021/09/boyalex/2405042847 head: Predict_large_cs_clever_clicks_2360931838_stand_mse

bert model 2021/09/boyalex/2405042847 head: Predict_large_google_2334139725_stand_mse

bert model 2021/09/boyalex/2405042847 head: Predict_pq_large_2394064809_stand_mse

bert model 2021/09/boyalex/2405042847 head: Predict_jul_xlarge_cs_sinsig_full_2394990987_stand_mse

bert model 2021/09/boyalex/2405042847 head: Predict_jul_xlarge_cs_apscore_2397363137_stand_mse

Count of news documents in PRS in top_1 document in all TMetaGroup's

Count of news documents in PRS

is query in popular queries list at 202108

DSSM prediction of google specificity for query