Tshuab xav muaj txoj kev los twb redefined peb lub neej nyob rau hauv ntau txoj kev. Ib tug piv txwv txog tej yam no yog li cas peb khoom nyob rau hauv kev yog thaum redefined. Thaum peb los ntawm khoom xauj, lub kaw lus pom zoo muab cov khoom uas peb yuav tau xav seb. Tsis hais tus foundations — ua hauj lwm los yog kev, Tau pom zoo lub nruab immensely lig. Thiab ntaub ntawv loj yog tsav hlab qab xav tshuab. Tus raug pom zoo uas tsis tau ua nws txoj hauj lwm tsis muaj ntaub ntawv txaus thiab loj cov khoom siv kom ntau neeg siv cov ntaub ntawv xws li tus muas khoom lawm, keeb kwm browsing, thiab tawm tswv yim rau tus xav tshuab los muab tswv yim zoo thiab ntau yam. Nyob rau hauv ib nutshell, Recommenders siab heev nws tsis tau tau zoo tsis muaj ntaub ntawv loj.
Lub kaw lus pom zoo ua hauj lwm li cas?
Lub kaw lus pom zoo ua haujlwm nyob rau hauv zoo hais tseg, zajlus kom theem uas yog tej ntaub ntawv, kev, thiab filtering. Cov theem uas nyob hauv qab.
Tej ntaub ntawv
Peb muaj kev pov plob hais tias ib tug neeg ntawm lub website Amazon browsing phau ntawv thiab nyeem kom paub meej. Txhua lub sij hawm tus nyeem ntawv clicks rau ib qhov txuas, raug rho yuav muaj xws li Ajax hem hem tawm haujlwm. Cov kev tshwm sim yam yuav sib txawv nyob ntawm seb siv lub tshuab. Cov kev tshwm sim ces yuav ua rau kev nkag teb chaws rau ib database uas feem ntau yog ib tug NoSQL database. Lub qhov rooj yog kev nyob hauv cov ntsiab lus tab sis tus sau lus txawj nyeem tej yam xws li "Neeg muaj khoom Z clicked lus ib zaug". Uas yog neeg siv cov lus kom yuav thiab zoo li cas muab cia rau yav tom ntej seb.
Lawv xav ntes kom paub meej npaum li cas? Yog hais tias tus neeg siv tau in thiaj li tso, ces cov lus no yog muab rho tawm xws li los ntawm kev sib kho http los yog ntawm lub ncuav qab zib uas. Nyob rau hauv cov ntaub ntawv pom zoo rau lawv nyob ntawm lawv ncuav qab zib, ces cov ntaub ntawv yeej muaj nyob rau xwb txog lub sij hawm tus neeg siv cov qub davhlau ya nyob twg. Koom kev kawm raug rho tawm haujlwm yuav luag nyob rau hauv txhua rooj plaub-ib tug neeg liking ib yam khoom los yog ntxiv rau lub laub nriav tsev thiab yuav. Yog li ntawd yog li cas txhawb neeg siv yuav muab. Tab sis, nws yog ib qhov chaw Recommenders ua li cas.
Cov paragraphs hauv qab no qhia li cas Amazon muaj qhov tswv yim nws cov khoom rau ib tug neeg uas yog browsing rau phau ntawv:
- Li ntawm daim duab hauv qab no, Thaum ib tug neeg nrhiav hauv phau Harry Potter thiab pob zeb 's Philosopher, ntau cov tswv yim kuj tau.
Lawv pom zoo
- Hauv lwm cov piv txwv, tus neeg uas raug tshawb Amazon rau Canon EOS 1200 D 18MP cov SLR koob yees duab (Dub nciab) interestingly tiv thaiv kab mob ntau tswv yim hauv cov koob yees duab accessories.
Lawv pom zoo
Kev
Kev tseem ceeb rau cov kev txiav txim zoo uas lawv qhia rau koj paub seb ib tug neeg pom tau tias txog ib yam khoom. Tus kev xav txog cov khoom ib yuav muab reflected kom ib tug twg nyob rau qhov nws yuav siv sij hawm xws li li nyiam, ntxiv mus tajlaj laub, kev yuav khoom los yog txhaj. Tshuab xav tau cob implicit kev raws li tej yam ua neeg. Qhov ntsuam xyuas ntau yog 5. Piv txwv, nriav tau yuav raug ua hauj lwm muaj kev ntsuam xyuas ntawm 4, yam koj nyiam ua tau 3, txhaj tau 2 thiab hais txog. Xav tshuab kuj yuav ua kev nyiaj thiab cov neeg siv tswv yim muab.
Filtering
Filtering filtering khoom raws li kev thiab lwm cov neeg siv cov ntaub ntawv hais. Tshuab xav siv peb txoj kev filtering: yaam, raws li neeg thiab ib cov kev hybrid. Nyob hauv koom filtering, ua ib tug sib piv ntawm cov neeg siv cov kev xaiv thiab cov tswv yim tiv thaiv kab mob. Piv txwv, Yog hais tias muaj neeg siv X nyiam khoom A, B, C, thiab tus D thiab cov neeg siv Y nyiam khoom A, B, C, D thiab E, tus nws hauv qhov ncauj dua tus neeg siv X yuav pub siv khoom E vim muaj ntau ntawm qhov sib thooj ntawm cov neeg siv X thiab Y kom xaiv cov khoom uas yog txhawj.
Ob peb reputed hom xws li Facebook, Twitter, LinkedIn, Amazon, Google News, Spotify thiab Last.fm siv cov qauv no muaj tswv yim zoo thiab ntau yam. Rau neeg raws li filtering, tus me nyuam twb browsing, nyiam, muas khoom thiab kev yuav tshwm sim rau hauv koj tus account ua ntej muab tswv yim. Cov qauv no yog siv los ntawm ntau reputed hom xws li IMDB, Txiv lws suav lwj thiab Pandora. Tuam txhab uas muag ntau dua li siv ib hybrid kev. Netflix tsis raug siv ib cov hybrid mus kom ze.
Luag hauj lwm uas muaj ntaub ntawv loj
Raws li tau teev lawm, cov ntaub ntawv loj drives Recommenders ua dab neeg. Recommenders tsis tau ua ib tug uas tsis muaj cov mov qhov ntawm cov ntaub ntawv. Txawm li cas los, lub luag hauj lwm ntawm cov ntaub ntawv loj tshaj xwb cov ntaub ntawv mus. Nws yog tseeb tias cov haujlwm hais los saum toj ntawd yuav tsum tau ib tug CPU muaj peev xwm uas yuav ua hauj lwm rau lub sij hawm. Kom paub qhov no, Yuav tsum siv Hadoop. Txo cov kev tuav hauj lwm ntxiv mus rau txoj, taw algorithms txoj, hau kev kev zoo sib thooj thiab lwm yam kev pab raws qib, Yuav tsum siv mahout.
Mahout yog lub tsev qiv ntawv uas comprises tshuab kawm algorithms. Nws qhia txog kev xaiv xaiv xav algorithm txheej, xaiv n-nyob ze tej zej zog thiab cov hau kev kev zoo sib thooj. Tab sis yog nws yog ib tug Java txheem kawm, nws koomtxoos rau Hadoop purely.
Kom koj paub tab txawm yooj yim, koj siv tau ib tug uas hu ua PredictionIO uas bundles cov Mahout thiab cov Hadoop thiab dab tsi ntau, nws muaj ib tug zoo neeg interface.
Li ntawd, lub luag hauj lwm ntawm cov ntaub ntawv loj yuav tau summed muab ntaus nqe hnyav, actionable cov ntaub ntawv ceev thiab muab tsim nyog teeb kom ceev ceev ntaub ntawv cov ntaub ntawv. Nws yog ib cuab kev tias nws yog tshuaj Hmoob yees yog tsis meant rau cov txheej txheem xws loj tagnrho cov ntaub ntawv no kom sai li sai. Li ntawd, nws yuav tsis suffice cia li muaj cov loj thiaj li yuav muab tswv yim muaj zog.
Amazon siv rooj plaub
Li cas Amazon siv cov haib duo loj cov ntaub ntawv thiab pom zoo lawv puas tsim nyog ib txoj kev tshawv no. Amazon nws kuj yog ib txoj kev uas ib pioneer ntawm ecommerce, tseem ceeb tshaj ntawd accolade yog li cas nws yog tus tsav tsheb cov nyiaj tau los ntawm kev muab tswv yim zoo ntau tshaj.
Muas kuj yog cov ua tsis nco xav thiab lawm thiab Amazon yog smartly tapping rau hauv txoj kev ua tsis nco xav shopper siab los ntawm kev muab tswv yim txog thiab pab tau khoom. Rau cov uas, nws relentlessly ua hauj lwm rau nws xav cav ua haib dua. Khw muaj ib qhov zoo sib xws rau psychology. Shoppers yuav rau instant gratification, PARASITE Instant uplift, kev esteem thiab lwm yam ntawd twb tsis paub tias lawv hais meej.
Amazon ntse txaus coj tej yam zoo rau hauv koj tus account yog. Thiab tam sim no, nws ua hauj lwm rau ib tug mob hu ua predictive dispatch uas txhais tau hais tias nws xav cav yuav twv seb dab tsi rov qab yuav tsum yuav thiab nrhiav kev pab ib tug dispatch hais.
Yog ua li cas thiaj Amazon tus achievements ntau creditable txog uas tsis zoo li Facebook — uas tseem relies heev rau tej ntaub ntawv loj — uas paub ntau paub meej txog nws cov subscribers, Amazon tag nrho uas paub txog cov neeg muas zaub yog cov qauv spending.
Amazon muaj tau pauv rau no paub smartly nyob rau hauv ib txoj kev los mus tshaj tawm ntawm koj twb. Nws yog ib txoj hauj lwm tsis yooj yim rau kev tsom xam cov qauv siv, nyiam, nyiam khoom thiab muaj tswv yim zoo rau tas li ntawd xwb. Thiab tam sim no, Amazon yog ua kom muaj cov cuab yeej thiab yees siv ntaub ntawv loj loj thiab xav tshuab zoo li ntawd, cia muag rau lwm corporations uas yog siv tej ntaub ntawv loj. Li ntawd, Amazon cov khoom tawm yuav pib feem ntau tshwm rau lwm cov websites thiab li zoo uas yuav tsav tau muag khoom.
Cov duab hauv qab no qhia seb lub tuam txhab uas muag yeej tau siv lub zog loj cov ntaub ntawv thiab pom zoo xyaw.
Tswv yim
Yus xav tshuab
Rau tag nrho lawv tshem, Xav Systems tsis muaj ib tug pov thawj uas puv tsim. Recommenders paub raug kev txom nyem los ntawm qhov cov kev txwv nram qab no:
- Recommenders vam lig rau tej ntaub ntawv thiab cov hirers yuav tsum lossi muab lawv loj tagnrho ntawm cov ntaub ntawv uas muaj. Yog vim li cas; me npe yog ntxiv raws tsis cuag luag ces lub npe loj xws li Google Amazon.
- Recommenders yuav pom tias nws nyuaj raws nraim qhia cov neeg siv yuav xaiv hais tias yog tus neeg nyiam yuav hloov sai, raws li thaum zam. Recommenders vam heev rau tej ntaub ntawv historic tab sis uas tej zaum yuav tsis haum rau tej yam khoom niches.
- Recommenders fim teeb meem nrog rau cov khoom unpredictable. Piv txwv, muaj ntau tej yam yeeb yam uas evoke huab tshuaj yog muaj xws li hlub los ntxub. Nws yog ib yam nyuaj heev los muab tswv yim rau tej yam khoom.
Txoj kev
Thaum cov ntaub ntawv loj loj thiab xav xyaw muaj twb proved ib thaum lawv tseem tsis tshua muaj neeg pab tau rau cov corporations loj loj, nws puas tsimnyog muaj txog seb cov tuam txhab uas muag nrog me pob nyiaj siv puas muaj nyiaj kev nqis peev uas zoo li no. Nws yog txhawb rau tej tuam txhab uas muag qhov cuab yeej ntawd cov ntaub ntawv loj loj thiab muaj yees kuj nyaj ntau. Seb khoom tseem ceeb heev thaum muaj ib tug zoo neeg siv kev ntawm seem yus pom tus neeg muas zaub. Tseem, ntawm lub tuam txhab seem yus pom, nws yuav siv mus ua nyiaj npaum li cas tiag factors uas yuav ua rau tus neeg uas yuav cov khoom uas yuav zoo li tsis khws pes tsawg. Raws li cov saum no duab qhia tau hais tias, lawm Recommenders zog loj.
