Tabulate cosine similarity/distance of word pairs.

Usage

tab_similarity(
  data,
  words = NULL,
  pattern = NULL,
  words1 = NULL,
  words2 = NULL,
  unique = FALSE,
  distance = FALSE
)

Arguments

data: A wordvec (data.table) or embed (matrix), see data_wordvec_load.
words: [Option 1] Character string(s).
pattern: [Option 2] Regular expression (see str_subset). If neither words nor pattern are specified (i.e., both are NULL), then all words in the data will be extracted.
words1, words2: [Option 3] Two sets of words for only n1 * n2 word pairs. See examples.
unique: Return unique word pairs (TRUE) or all pairs with duplicates (FALSE; default).
distance: Compute cosine distance instead? Defaults to FALSE (cosine similarity).

Value

A data.table of words, word pairs, and their cosine similarity (cos_sim) or cosine distance (cos_dist).

Download

Download pre-trained word vectors data (.RData): https://psychbruce.github.io/WordVector_RData.pdf

Examples

tab_similarity(demodata, cc("king, queen, man, woman"))
#>     word1 word2    wordpair   cos_sim
#>  1:  king  king   king-king 1.0000000
#>  2:  king queen  king-queen 0.6510958
#>  3:  king   man    king-man 0.2294268
#>  4:  king woman  king-woman 0.1284797
#>  5: queen  king  queen-king 0.6510958
#>  6: queen queen queen-queen 1.0000000
#>  7: queen   man   queen-man 0.1665822
#>  8: queen woman queen-woman 0.3161813
#>  9:   man  king    man-king 0.2294268
#> 10:   man queen   man-queen 0.1665822
#> 11:   man   man     man-man 1.0000000
#> 12:   man woman   man-woman 0.7664012
#> 13: woman  king  woman-king 0.1284797
#> 14: woman queen woman-queen 0.3161813
#> 15: woman   man   woman-man 0.7664012
#> 16: woman woman woman-woman 1.0000000
tab_similarity(demodata, cc("king, queen, man, woman"),
               unique=TRUE)
#>    word1 word2    wordpair   cos_sim
#> 1:  king queen  king-queen 0.6510958
#> 2:  king   man    king-man 0.2294268
#> 3:  king woman  king-woman 0.1284797
#> 4: queen   man   queen-man 0.1665822
#> 5: queen woman queen-woman 0.3161813
#> 6:   man woman   man-woman 0.7664012

tab_similarity(demodata, cc("Beijing, China, Tokyo, Japan"))
#>       word1   word2        wordpair   cos_sim
#>  1: Beijing Beijing Beijing-Beijing 1.0000000
#>  2: Beijing   China   Beijing-China 0.7648461
#>  3: Beijing   Tokyo   Beijing-Tokyo 0.5229628
#>  4: Beijing   Japan   Beijing-Japan 0.3995245
#>  5:   China Beijing   China-Beijing 0.7648461
#>  6:   China   China     China-China 1.0000000
#>  7:   China   Tokyo     China-Tokyo 0.3814305
#>  8:   China   Japan     China-Japan 0.5967756
#>  9:   Tokyo Beijing   Tokyo-Beijing 0.5229628
#> 10:   Tokyo   China     Tokyo-China 0.3814305
#> 11:   Tokyo   Tokyo     Tokyo-Tokyo 1.0000000
#> 12:   Tokyo   Japan     Tokyo-Japan 0.7002254
#> 13:   Japan Beijing   Japan-Beijing 0.3995245
#> 14:   Japan   China     Japan-China 0.5967756
#> 15:   Japan   Tokyo     Japan-Tokyo 0.7002254
#> 16:   Japan   Japan     Japan-Japan 1.0000000
tab_similarity(demodata, cc("Beijing, China, Tokyo, Japan"),
               unique=TRUE)
#>      word1 word2      wordpair   cos_sim
#> 1: Beijing China Beijing-China 0.7648461
#> 2: Beijing Tokyo Beijing-Tokyo 0.5229628
#> 3: Beijing Japan Beijing-Japan 0.3995245
#> 4:   China Tokyo   China-Tokyo 0.3814305
#> 5:   China Japan   China-Japan 0.5967756
#> 6:   Tokyo Japan   Tokyo-Japan 0.7002254

## only n1 * n2 word pairs across two sets of words
tab_similarity(demodata,
               words1=cc("king, queen, King, Queen"),
               words2=cc("man, woman"))
#>    word1 word2    wordpair    cos_sim
#> 1:  king   man    king-man 0.22942676
#> 2:  king woman  king-woman 0.12847968
#> 3: queen   man   queen-man 0.16658216
#> 4: queen woman queen-woman 0.31618132
#> 5:  King   man    King-man 0.15777646
#> 6:  King woman  King-woman 0.06369529
#> 7: Queen   man   Queen-man 0.09365463
#> 8: Queen woman Queen-woman 0.20171619