Tsvaga kuti matanho anobatsira sei kuchengeta bhokisi rako remukati rakachena
Bayesian spam filters vanoverenga mikana yemashoko ari spam kubva pane zviri mukati. Kusiyana nemafungiro akareruka-akachengetedzwa, Bayesian spam filtering inodzidza kubva kune spam uye kubva kumamishinari akanaka, zvichiita kuti pave nehutano hwakasimba, huchigadzirisa uye huri nani hunopesana nespam, iyo inonyanya kukosha, inodzoka pasina chero ipi zvayo yenhema.
Unoziva Sei Junk Email?
Funga nezvekuti unoona sei spam . Kukurumidza kuona kunowanzokwanira. Iwe unoziva kuti spam inotaridzika sei, uye iwe unoziva kuti mail yakanaka yakaita sei.
Iyo mikana ye spam inotarisa semakadhi akanaka yakapoteredza ... zero.
Kugovera Zvakagadzirirwa-Zvakagadzirwa Mafirimu Usasarudza
Hazvingave zvakanaka here kana zvipfeko zve spam zvinowanzoshanda zvakadaro, zvakare?
Kugovera zvigadziro-based based spam filters vanoedza izvozvo chete. Vanotsvaka mazwi nemamwe maitiro anowanzoitwa spam. Zvinhu zvose zvinoratidzirwa zvinopiwa zvikwereti, uye chikamu chepamu chemashoko ose chinotorwa kubva kune mumwe munhu zvidzidzo. Vamwe vanongoridza mafiritsi zvakare vanotarisa maitiro edzimisi yakakodzera, kuderedza chikamu chekupedzisira chemashoko.
Nzira yekutsvaga mafikisi inoshandiswa inoshanda, asi inewo ine zvipingamupinyi zvakawanda:
- Nhamba yezvigadziro inovakwa kubva kune spam (uye tsamba yakanaka) inowanikwa kune injinjini yefaira. Kuti uwane ruzivo rwakanaka rwechimiro che spam chero munhu upi zvake aigona kuwana, mail inofanira kuunganidzwa pamakero e-email. Izvi zvinowedzera kushanda kwei filters, kunyanya nokuti maitiro emashoko akanaka achave akasiyana kune mumwe nomumwe , asi izvi hazvifungi.
- Zviratidzo zvaunotarisa ndezvishoma kana zvishoma zvakagadzirirwa mumatombo . Kana vese spammers vanoedza kuchinja (uye kuita kuti spam yavo ione seyamasa yakanaka kune mafungiro), kuchena kunofanirwa kushandiswa nemaune-chinhu chakakura zvikuru.
- Nharo dzakagoverwa kune rimwe nerimwe shoko dzinogona kunge dzichibva pakufungidzira kwakanaka, asi richiri kushandura. Uye sehuwandu hwezviratidzo, hazvibvumirane nemamiriro ekushandura kwenyika ye spam mune zvekare kana kune zvinodiwa nemumwe munhu.
Bayesian Spam Filters Tweak Ivo pachavo, Kuwana zviri nani uye zviri nani
Bayesian spam filters ndeyemhando yekuongororwa-content-based filters, zvakare. Nzira yavo inopedza matambudziko ezvinyorwa zvinyorwa zvinyorwa zve spam, kunyange zvakadaro, uye zvinodaro zvakanyanya. Kubva kushaya simba kwekuisa zvikwangwani zviri muzvinyorwa zvakagadzirirwa maitiro uye mazana avo, urongwa urwu huchabviswa.
Pane kudaro, vaBasesian spam filtara vanovaka urongwa ivo pachavo. Zvakanaka, iwe unotanga ne (guru) boka remaimeri rawakaronga se spam, uye rimwe boka remashoko akanaka. Aya mafungiro anotarisa zvose uye akaongorora mail yakakodzera pamwe ne spam kuti aone mukana wezvinhu zvakasiyana-siyana zvinowanikwa mu spam, uye mumakadhi akanaka.
Kuti Bayesian Spam Filter Inoongorora sei Imeyili
Zvinhu zvinoshandiswa neBasesian spam filter zvinogona kutarisa zvinogona kuva:
- mashoko mumutumbi weshoko, hongu, uye
- zvinyorwa zvaro (vatumwa nemigwagwa dzemigwagwa , somuenzaniso!), asiwo
- zvimwe zvinhu zvakadai se HTML / CSS code (semirau uye mamwe maitiro), kana kunyange
- shoko pairi, mabheji uye
- meta yeruzivo (apo mumwe mutauro unowanikwa, somuenzaniso).
Kana izwi, "Cartesian" somuenzaniso, risingambooneki mu spam asi kazhinji mumavhurairi akakodzera waunowana, mukana wekuti "Cartesian" inoratidza spam iri pedyo nezero. "Toner", kune rumwe rutivi, inowanikwa chete, uye kazhinji, mu spam. "Toner" ine mikana yakakwirira yekuwanikwa mu spam, kwete zhinji pasi pe1 (100%).
Apo shoko idzva rinosvika, rinongororwa neBazesian spam filter, uye mikana yemashoko akazara ari spam inotorwa achishandiswa munhu mumwe chete maitiro.
Funga kuti mharidzo inosanganisira zvose "Cartesian" uye "toner". Kubva pamashoko aya chete haasati ajekesa kana tine spam kana mail yakarurama. Dzimwe humwe huchaita (zvinotarisirwa uye zvichida zvichida) zvinoratidza chiitiko chinobvumira firiji kuti iparidze shoko seye spam kana mail yakanaka.
Bayesian Spam Filters Inogona Kudzidza Kushandura
Iye zvino kuti tine chikamu, shoko rinogona kushandiswa kudzidzisa firisi yacho pachayo zvakare. Muchiitiko ichi, kana mukana we "Cartesian" unoratidza mutsamba wakanaka unoderedzwa (kana shoko rine zvose "Cartesian" uye "toner" rinowanikwa se spam), kana kuti mikana ye "toner" inoratidza kuti spam inofanira kuongororwa.
Kushandisa nzira iyi inogadziriswa, Bayesian filters vanogona kudzidza kubva kumadziro avo pachavo uye kusarudza kwevashumiri (kana iye achigadzirisa kusaruramisirwa nemafuta). Kugadziriswa kweBayesian kuongorora kunowedzerawo kuti vanobudirira zvikuru kune mumwe munhu weimeri. Kunyange zvazvo vanhu vazhinji ve spam vangave nemafungiro akafanana, kero yakakodzera yakasiyana neyevanhu vose.
Ndezvipi Zvinogona Kuparadzaniswa Zvichida Kuwana Mafiriji eBaseesian?
Mamiriro edzimisi yakakodzera yakakosha zvikuru kuBasesian spam filtering process sezvo spam iri. Kana ma filters akadzidziswa zvakananga kumushandi wese, spammers achava nenguva yakaoma zvikuru kushanda kumativi ose evanhu (kana kuti kunyange vanhu vazhinji) filters spam, uye mafirita anogona kuenderana nezvose zvinotengesa spammers.
Spammers inongoita kuti ipfuure zvakanyatsodzidziswa mabhesi eBhaesiesi kana vakaita sarudzo dzavo dzepa spam dzichitarisa zvakakwana seyemajeri evanhu vose vanogona kuwana.
Spammers havawanzo kutumira ma-mail akadai. Ngatitorei izvi ndezvokuti idzi emaimeri hazvishandi se email isina basa. Saka, mukana iwoyo haazove achiita apo mavara emarenda, anonyadzisa ndiyo nzira chete yekuita kuti ipfuure zvakapera filer.
Kana spammers ichishandura kune ema-mail akawanda anotarisa, zvisinei, tichaona yakawanda spam muInboxes yedu zvakare, uye email inogona kunge inogumbura sezvayakanga iri mumazuva ekutanga eBasesian (kana kunyanya kuipa). Izvo zvichange zvakaparadza mumusika wemhando dzakawanda dze spam, zvisinei, uye naizvozvo hazvizoperi kwenguva refu.
Nzira Dzakasimba Dzinogona kuva Bayesian Spam Filter & Achi; # Achilles & # 39; Heel
Chimwe chete chinogona kuonekwa kuti spammers kuti vashandise nzira dzavo kuburikidza nemafuta eBazesian kunyange neavo vanowanzogutsikana. Icho chiri mumamiriro ekuverengwa kweBazesian kuti shoko rimwe kana chimiro chinonyanya kuwanikwa mumachesi akanaka chinogona kuva chakakosha sekutsvaga chero mharidzo kubva pakutarisa se spam kuti ienzaniswe se ham nefriji.
Kana spammers vachiwana nzira yekuziva yako chokwadi-moto moto ma-mail-kuburikidza nekushandisa maRIT receipt kuti uone marimiro awakavhura, somuenzaniso-, anogona kusanganisira mumwe wavo mumisipiresi isina kutora uye kusvika kwauri kuburikidza nepamusoro- akadzidziswa Bayesian filter.
John Graham-Cumming akaedza izvi nekurega mabheji maviri eBazesian achishandirana, "yakaipa" iyo inogadzirisa kune mimwe mitauro inowanikwa kuti iwane mu "yakanaka". Anoti inoshanda, asi iyo inopedza nguva uye yakaoma. Hatifungi kuti tichaona zvakawanda zvezvizvi zvichiitika, zvisingaiti pamwero mukuru, uye hazvifananidzi kune ma email evanhu. Spammers inogona (kuedza) kuverenga mamwe mazwi makuru emasangano (chimwe chinhu se "Almaden" kune vamwe vanhu paBM zvichida?) Panzvimbo.
Kazhinji, spam ichagara iine (zvakakosha) yakasiyana neyokutumira tsamba kana kuti ichange isina spam, zvakadaro.
Icho Chinyorwa: Bayesian Filtering & # 39; s Simba Inogona Kuve Kushaya Kwayo
Bayesian spam filters zvakagutsikana-based filters izvo:
- vakanyatsodzidziswa kuti vaone mumwe munhu weEmam user's spam uye ma mail akanaka , achiita kuti zvibudirire kwazvo uye zvakaoma kuchinja kune spammers.
- rinogona kuramba uye risina basa rakawanda kana kuongorora mabhuku kunoshandiswa kune maitiro ekupedzisira e spammers.
- tora mail yega yega yega yega uye tive nehuwandu hwakanyanya hwehuwandu hwenhema .
- Zvinosuruvarisa, kana izvi zvichitadzisa kusavimbika muBasesian anti-spam filters, inoita kuti chikanganiso ichocho chiwedzere kunyanya . Zvimwe zvinopesana nezvakaipa zvenhema (spam inotarisa zvakananga sekutumira tsamba nguva dzose) ine simba rekuvhiringidza uye kuvhiringidza vashandisi.