{"id":372971,"date":"2024-05-21T05:25:10","date_gmt":"2024-05-21T05:25:10","guid":{"rendered":"http:\/\/savepearlharbor.com\/?p=372971"},"modified":"-0001-11-30T00:00:00","modified_gmt":"-0001-11-29T21:00:00","slug":"","status":"publish","type":"post","link":"https:\/\/savepearlharbor.com\/?p=372971","title":{"rendered":"\u0410\u043b\u0433\u043e\u0440\u0438\u0442\u043c \u0440\u0430\u0441\u0447\u0451\u0442\u0430 \u0440\u0430\u0441\u0441\u0442\u043e\u044f\u043d\u0438\u044f \u043c\u0435\u0436\u0434\u0443 \u0441\u0442\u0440\u043e\u043a\u0430\u043c\u0438<\/span>"},"content":{"rendered":"
<\/div>\n
\n
\n
\n
\n

\u041f\u043e \u0440\u0430\u0431\u043e\u0442\u0435 \u0441\u0442\u043e\u044f\u043b\u0430 \u0437\u0430\u0434\u0430\u0447\u0430 \u043e\u043f\u0442\u0438\u043c\u0438\u0437\u0430\u0446\u0438\u0438 \u043f\u043e\u0438\u0441\u043a\u0430 \u043f\u043e \u0430\u0434\u0440\u0435\u0441\u0430\u043c (\u0443\u043b\u0438\u0446\u044b, \u0434\u043e\u043c\u0430 \u0438 \u043e\u0431\u044a\u0435\u043a\u0442\u044b). \u0413\u043b\u0430\u0432\u043d\u044b\u0439 \u043a\u0440\u0438\u0442\u0435\u0440\u0438\u0439 — \u043d\u0430\u0445\u043e\u0436\u0434\u0435\u043d\u0438\u0435 \u0430\u0434\u0440\u0435\u0441\u0430, \u0435\u0441\u043b\u0438 \u043d\u0430\u043f\u0438\u0441\u0430\u043d\u043e \u0441 \u043e\u0448\u0438\u0431\u043a\u0430\u043c\u0438 \u0438\u043b\u0438 \u043d\u0435 \u0434\u043e\u043f\u0438\u0441\u0430\u043d \u043e\u043d \u0432 \u043f\u043e\u043b\u043d\u043e\u0439 \u043c\u0435\u0440\u0435. Bert\u2019\u044b, \u043a\u043e\u0441\u0438\u043d\u0443\u0441\u043d\u044b\u0435 \u0440\u0430\u0441\u0441\u0442\u043e\u044f\u043d\u0438\u044f \u044d\u043c\u0431\u0435\u0434\u0434\u0438\u043d\u0433\u043e\u0432 \u0438 \u0442.\u0434. \u043d\u0435 \u043f\u043e\u0434\u0445\u043e\u0434\u0438\u043b\u0438, \u0442\u0430\u043a \u043a\u0430\u043a \u043e\u043d\u0438 \u0437\u0430\u0442\u043e\u0447\u0435\u043d\u044b \u043f\u043e\u0434 \u0441\u043c\u044b\u0441\u043b\u043e\u0432\u043e\u0439 \u043f\u043e\u0438\u0441\u043a, \u0430 \u0432 \u0430\u0434\u0440\u0435\u0441\u0430\u0445 \u0441\u043c\u044b\u0441\u043b\u0430 \u043d\u0435\u0442. TF-IDF c \u043b\u0435\u043c\u043c\u0430\u0442\u0438\u0437\u0430\u0446\u0438\u0435\u0439 \u0442\u043e\u0436\u0435 \u043d\u0435 \u043e\u0447\u0435\u043d\u044c \u043f\u043e\u0434\u0445\u043e\u0434\u0438\u043b \u0434\u043b\u044f \u044d\u0442\u043e\u0439 \u0437\u0430\u0434\u0430\u0447\u0438, \u0440\u0435\u0437\u0443\u043b\u044c\u0442\u0430\u0442\u044b \u0431\u044b\u043b\u0438 \u043f\u043b\u043e\u0445\u0438\u0435.<\/p>\n

\u0414\u043b\u044f \u0440\u0435\u0430\u043b\u0438\u0437\u0430\u0446\u0438\u0438 \u043d\u0430\u0447\u0430\u043b \u0438\u0441\u043f\u043e\u043b\u044c\u0437\u043e\u0432\u0430\u0442\u044c \u0440\u0430\u0441\u0441\u0442\u043e\u044f\u043d\u0438\u0435 \u0414\u0430\u043c\u0435\u0440\u0430\u0443-\u041b\u0435\u0432\u0435\u043d\u0448\u0442\u0435\u0439\u043d\u0430<\/a>, \u0438 \u0432 \u043f\u043e\u0441\u043b\u0435\u0434\u0441\u0442\u0432\u0438\u0435, \u0440\u0430\u0437\u0432\u0438\u043b \u044d\u0442\u043e \u0434\u043e \u0441\u043e\u0431\u0441\u0442\u0432\u0435\u043d\u043d\u043e\u0433\u043e \u0430\u043b\u0433\u043e\u0440\u0438\u0442\u043c\u0430, \u043a\u043e\u0442\u043e\u0440\u044b\u0439 \u043d\u0430\u0445\u043e\u0434\u0438\u0442 \u0440\u0430\u0441\u0441\u0442\u043e\u044f\u043d\u0438\u0435 \u043c\u0435\u0436\u0434\u0443 \u0434\u0432\u0443\u043c\u044f \u0441\u0442\u0440\u043e\u043a\u0430\u043c\u0438.<\/p>\n

\u0426\u0435\u043b\u044c \u0434\u0430\u043d\u043d\u043e\u0433\u043e \u043f\u043e\u0441\u0442\u0430 \u043e\u043f\u0438\u0441\u0430\u043d\u0438\u0435 \u0442\u043e\u043b\u044c\u043a\u043e \u0430\u043b\u0433\u043e\u0440\u0438\u0442\u043c\u0430.<\/p>\n

\u041e\u043f\u0438\u0441\u0430\u043d\u0438\u0435 \u0430\u043b\u0433\u043e\u0440\u0438\u0442\u043c\u0430<\/h2>\n

\u041f\u0440\u043e\u0446\u0435\u0441\u0441 \u0440\u0430\u0431\u043e\u0442\u044b \u0430\u043b\u0433\u043e\u0440\u0438\u0442\u043c\u0430<\/h4>\n
<\/figure>\n

\u0424\u043e\u0440\u043c\u0443\u043b\u0430<\/h4>\n

Tew — \u044d\u0442\u043e \u043f\u043e\u043b\u0443\u0447\u0430\u0435\u043c\u043e\u0435 \u0440\u0430\u0441\u0441\u0442\u043e\u044f\u043d\u0438\u0435 \u043c\u0435\u0436\u0434\u0443 \u0434\u0432\u0443\u043c\u044f \u0441\u0442\u0440\u043e\u043a\u0430\u043c\u0438.<\/p>\n

<\/figure>\n

\u0415\u0441\u043b\u0438 \u0440\u0430\u0441\u0441\u043c\u043e\u0442\u0440\u0435\u0442\u044c \u043f\u043e \u044d\u0442\u0430\u043f\u0430\u043c:<\/p>\n

<\/figure>\n

\u0420\u0435\u0430\u043b\u0438\u0437\u0430\u0446\u0438\u044f \u043a\u043e\u0434\u043e\u043c \u043d\u0430 R<\/h2>\n

\u0424\u0443\u043d\u043a\u0446\u0438\u044f \u0440\u0435\u0430\u043b\u0438\u0437\u0430\u0446\u0438\u0438 \u0430\u043b\u0433\u043e\u0440\u0438\u0442\u043c\u0430.<\/p>\n

library(stringr) library(data.table) library(magrittr) library(stringdist)  \u041f\u0440\u0438\u0441\u043e\u0435\u0434\u0438\u043d\u044f\u044e \u043f\u0430\u043a\u0435\u0442: 'stringdist' \u0421\u043b\u0435\u0434\u0443\u044e\u0449\u0438\u0439 \u043e\u0431\u044a\u0435\u043a\u0442 \u0441\u043a\u0440\u044b\u0442 \u043e\u0442 'package:magrittr':      extract stringDistination <- function(src_query, trg_query){   clearSplit <- function(x){     x %>%        strsplit(split = \" \") %>%        unlist() %>%        tolower() %>%        .[. %ilike% \"\\\\w+\"] %>%        str_squish()   }    # Splitting queries   src = clearSplit(src_query)   trg = clearSplit(trg_query)         # By every word find closeness   x <- list()   y <- c()   for (i in seq(length(src))) {     for (l in seq(length(trg))) {       y[trg[l]] <- stringdist(a = trg[l], b = src[i], weight =  c(0.8, 1, 1, 1)) \/ max(nchar(src[i]), nchar(trg[l]), na.rm = TRUE) # finding similarity word by word     }     x[[src[i]]] <- y   }      # min value by each word   vec <- x %>%      unlist() %>%      {       str <- names(.) %>% tstrsplit(split = \"\\\\.\")       c(         setNames(           object = .,           nm = str[[1]]         ),         setNames(           object = .,           nm = str[[2]]         )       )     } %>%     sort %>%      .[unique(names(.))]      # finding digitals   num_names <- which(str_detect(string = names(vec), pattern = \"\\\\d+\"))      # if number, number have to be low impactable   if(length(num_names) > 0){     mean_val <- mean(vec, na.rm = TRUE)     vec[num_names] <- fifelse(vec[num_names] > mean_val, vec[num_names], mean_val)   }      vec %>%      .[seq(length(src))] %>%     {       . * (1\/log(order(.)+1)) # calculate distance value depending on its position     } %>%     mean() %>%  # take mean value     {       if(length(src) == length(trg)){ . * (1 - 0.005)} else {.}  # if lengths of src and trg are equal, in this case the obtained average value is reduced by 0.5%     } }<\/code><\/pre>\n
\u041f\u0440\u0438\u043c\u0435\u0440 \u0440\u0430\u0431\u043e\u0442\u044b \u0444\u0443\u043d\u043a\u0446\u0438\u0438<\/h4>\n
queries <- c(\"\u042d\u0441\u043a\u0438 \u0441\u0430\u0440\u044b\", \"\u041d\u0430\u0440\u0442\u044b\") target_vector <- c(\"\u042d\u0441\u043a\u0438 \u0441\u0430\u0440\u044b \u043a\u0451\u043b\", \"\u0425\u0430\u0441\u0430\u0443\u0442\u0441\u043a\u0430\u044f\", \"\u041d\u0430\u0440\u0442\u043e\u0432\", \"\u041d\u043e\u0432\u044b\u0439 \u041a\u0430\u0440\u0430\u0447\u0430\u0439\", \"\u041c\u0430\u0440\u0430-\u0410\u044f\u0433\u044a\u044b\", \"\u041a\u0430\u0432\u043a\u0430\u0437\u0441\u043a\u0430\u044f\")  data.table(query = rep(queries, each = length(target_vector)),            target = rep(target_vector, length(queries))) %>%    .[, similarity := stringDistination(src_query = query, trg_query = target) %>% round(3), by = 1:nrow(.)] %>%    .[, .SD[order(similarity)], by = query] %>%    .[, position := 1:.N, by = query]          query        target similarity position 1  \u042d\u0441\u043a\u0438 \u0441\u0430\u0440\u044b \u042d\u0441\u043a\u0438 \u0441\u0430\u0440\u044b \u043a\u0451\u043b      0.000        1 2  \u042d\u0441\u043a\u0438 \u0441\u0430\u0440\u044b        \u041d\u0430\u0440\u0442\u043e\u0432      0.784        2 3  \u042d\u0441\u043a\u0438 \u0441\u0430\u0440\u044b    \u041c\u0430\u0440\u0430-\u0410\u044f\u0433\u044a\u044b      0.824        3 4  \u042d\u0441\u043a\u0438 \u0441\u0430\u0440\u044b \u041d\u043e\u0432\u044b\u0439 \u041a\u0430\u0440\u0430\u0447\u0430\u0439      0.836        4 5  \u042d\u0441\u043a\u0438 \u0441\u0430\u0440\u044b    \u0425\u0430\u0441\u0430\u0443\u0442\u0441\u043a\u0430\u044f      0.941        5 6  \u042d\u0441\u043a\u0438 \u0441\u0430\u0440\u044b    \u041a\u0430\u0432\u043a\u0430\u0437\u0441\u043a\u0430\u044f      0.941        6 7      \u041d\u0430\u0440\u0442\u044b        \u041d\u0430\u0440\u0442\u043e\u0432      0.478        1 8      \u041d\u0430\u0440\u0442\u044b \u042d\u0441\u043a\u0438 \u0441\u0430\u0440\u044b \u043a\u0451\u043b      0.519        2 9      \u041d\u0430\u0440\u0442\u044b    \u041c\u0430\u0440\u0430-\u0410\u044f\u0433\u044a\u044b      1.005        3 10     \u041d\u0430\u0440\u0442\u044b \u041d\u043e\u0432\u044b\u0439 \u041a\u0430\u0440\u0430\u0447\u0430\u0439      1.030        4 11     \u041d\u0430\u0440\u0442\u044b    \u0425\u0430\u0441\u0430\u0443\u0442\u0441\u043a\u0430\u044f      1.148        5 12     \u041d\u0430\u0440\u0442\u044b    \u041a\u0430\u0432\u043a\u0430\u0437\u0441\u043a\u0430\u044f      1.292        6 <\/code><\/pre>\n
\u0412\u044b\u0432\u043e\u0434<\/h2>\n
\u041f\u043e \u0441\u0440\u0430\u0432\u043d\u0435\u043d\u0438\u044e \u0441 \u0441\u0443\u0449\u0435\u0441\u0442\u0432\u0443\u044e\u0449\u0438\u043c\u0438 \u043f\u043e\u0438\u0441\u043a\u043e\u0432\u044b\u043c\u0438 \u0434\u0432\u0438\u0436\u043a\u0430\u043c\u0438 (\u0430 \u0441\u0440\u0430\u0432\u043d\u0438\u0432\u0430\u043b \u0441 meilisearch) \u0435\u0441\u0442\u044c \u0441\u0432\u043e\u0438 \u043f\u0440\u0435\u0438\u043c\u0443\u0449\u0435\u0441\u0442\u0432\u0430:<\/p>\n