BM25 na Mweghachi Lexical
BM25 bụ ọrụ nrịbama okwu gbadoro ụkwụ n'okwu na-agụ akwụkwọ site na ugboro ole okwu ajụjụ na-apụta, gbanwee maka ụkọ okwu na ogologo akwụkwọ.
Nchịkọta
Decades old, it remains a remarkably strong and ubiquitous baseline for search.
Ime miri emi
BM25 (Best Matching 25) bụ ọrụ ogo akpa-okwu sitere na usoro Okapi puru omume nke 1990s. Maka okwu ajụjụ ọ bụla ọ na-ejikọta akara atọ: ugboro ole okwu ahụ pụtara ( ugboro ole okwu ahụ na-apụta na akwụkwọ, yana nloghachi na-ebelata nke oke k1 na-achịkwa), ugboro ole akwụkwọ na-agbanwe agbanwe (okwu ndị dị ụkọ n'ofe mkpokọta na-agụ karịa), na nhazi ogologo akwụkwọ (parameter b, ya mere akwụkwọ ogologo na-adịghị mma). Kpọkọta ọnụ ọgụgụ ndị a n'otu oge wee nweta ọkwa akwụkwọ ahụ. Ọ dịghị achọ ọzụzụ ma na-agba ọsọ ngwa ngwa site na ntụgharị ntụgharị, nke mere engines ọchụchọ dị ka Elasticsearch na Lucene jiri ya na ndabara. N'agbanyeghị ịrị elu nke iweghachite akwara ozi, BM25 ka na-emeri ma ọ bụ njikọ n'ọtụtụ akara ngosi, ọkachasị maka okwu ndị na-adịghị ahụkebe, ihe nchọpụta ziri ezi, na ajụjụ ndị na-abụghị ngalaba.
Nghọta nka nka
Akụkụ ugboro ugboro BM25 na-ejupụta: oke k1 na-ekpebi ole okwu ugboro ugboro na-ebuli akara, ya mere okwu pụtara ugboro 50 abụghị 50x dị mkpa karịa otu ugboro. Paramita b na-agwakọta oge raw na ogologo-normalized ugboro. IDF na-ewetu okwu nkịtị dị ka 'nke' ma na-enye ndị pụrụ iche ụgwọ ọrụ. N'ihi na ọ na-arụ ọrụ na ndeksi ntụgharị ntụgharị, na-edobe mkpụrụokwu ọ bụla na ndepụta akwụkwọ ya, akara akara na-emetụ naanị akwụkwọ nwere okwu ajụjụ, na-eme ka ọ rụọ ọrụ nke ọma.
Mmetụta atụmatụ
Ọsọ na ọnụ ọgụgụ
Usoro ọrụ asụsụ nwere ike ịga ngwa ngwa n'achụghị nkwụsi ike.
Nweta na iru
Ọ na-agbasawanye ohere n'ofe asụsụ na ụdị nzikọrịta ozi.
Mkpebi doro anya
Ndị otu nwere ike itinyekwu oge na ikpe ebe akpaaka na-ejikwa nkwughachi.
Ọdịnihu nke BM25 na Mweghachi Lexical
BM25 eyighị ka ọ ga-apụ n'anya; kama ọ na-esiwanye ike na usoro akwara ozi n'ịweghachi ngwakọ, ebe a na-ejikọta ọnụ ọgụgụ lexical na oke (na-abụkarị site n'usoro ngbanwe). Modelsdị obere amụtara dị ka SPLADE na-agwakọta sparsity ụdị BM25 yana ịdị arọ okwu akwara, yana BM25 na-arụkarị ọrụ dị ka onye na-eweghachite ọkwa mbụ n'ihu ndị na-emegharị akwara. Ọsọ ya, nkọwa ya na ọnụ ahịa ọzụzụ efu na-ekwe nkwa ọrụ na-adịgide adịgide n'ịchọ mmepụta.
Mmejuputa n'ezie n'ụwa
Ọkwa dị mkpa na Elasticsearch, OpenSearch, na Apache Lucene/Solr
Mweghachite onye ọkwa ọkwa mbụ nke na-enye onye na-emegharị akwara ji nwayọ nwayọ na nchọ ọkwa abụọ
Koodu na nchọta ndekọ ebe ihe njirimara na koodu njehie ga-adakọrịrịrịrị
Na-egwupụta ihe atụ na-adịghị mma iji zụọ ndị na-eweghachite oke dị ka DPR
Ihe ize ndụ & okporo ụzọ nche
Eziokwu ndị e chepụtara echepụta nwere ike jiri nwayọ tinye akụkọ, nkwado nkwado, ma ọ bụ nsonaazụ nyocha.
Mmetụta ngwa ngwa nwere ike ịmepụta nsonaazụ na-ekwekọghị ekwekọ n'ofe arịrịọ ndị yiri ya.
Enwere ike ikpughe data ederede nwere mmetụta ma ọ bụrụ na njikwa ohere adịghị ike.
Map mmejuputa
Kọwaa usoro mmepụta, ụda, na ụkpụrụ ịdịmma tupu ibugharị.
Weghachite nzaghachi site na isi mmalite ntụkwasị obi mgbe ọ bụla izi ezi dị mkpa.
Debe ebe nleba anya mmadụ maka mpụta dị elu.
Sochie ụkpụrụ ọdịda ma na-azụghachi mkpali ma ọ bụ usoro ọrụ mgbe niile.
Nọgide na-eme nchọpụta
Free newsletter
Get the daily AI briefing
Three verified AI stories every weekday morning, written in plain English. Free forever, no ads.
One email each weekday. Unsubscribe in one click. We never sell or share your address.
Test yourself
Take the BM25 and Lexical Retrieval quiz
Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.
Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation
Ntuziaka na-esote
Mweghachite mmekọrịta ikpeazụ nke ColBERT
Ajụjụ a na-ajụkarị
What is BM25 and Lexical Retrieval?
BM25 bụ ọrụ nrịbama okwu gbadoro ụkwụ n'okwu na-agụ akwụkwọ site na ugboro ole okwu ajụjụ na-apụta, gbanwee maka ụkọ okwu na ogologo akwụkwọ. Afọ iri gara aga, ọ na-anọgide na-abụ ntọala siri ike na ebe niile maka ịchọ.
Kedu ihe BM25 na-eji eme ihe n'ọkwa akwụkwọ?
BM25 na-ejikọta ugboro okwu, ugboro ole akwụkwọ na-agbanwe agbanwe (nke ụkọ okwu), na nhazi ogologo akwụkwọ ka ọ bụrụ akara mkpa.
Kedu ọrụ ugboro ugboro akwụkwọ inverse (IDF) na-arụ na BM25?
IDF na-akwụ ụgwọ okwu ndị na-adịghị ahụkebe n'ofe mkpokọta ahụ ma na-agbadata okwu nkịtị, ebe ọ bụ na egwuregwu na-adịghị ahụkebe na-enye ihe ọmụma.
Kedu ihe kpatara BM25 ji etinye normalization ogologo akwụkwọ (ihe b parameter)?
Na-enweghị normalization, akwụkwọ ogologo ga-akwakọba ọtụtụ okwu ọkụ; paramita b na-edozi ogologo ya ka ntụnyere dị mma.
Kedu usoro data na-eme BM25 ngwa ngwa n'ọtụtụ?
Ndekọ ntụgharị ntụgharị na-eme ka BM25 nweta akara naanị akwụkwọ nwere okwu ajụjụ, na-eme ka iweghachite dị ezigbo mma.