quanteda has the functionality to select, remove or compound multi-word expressions, such as phrasal verbs (“try on”, “wake up” etc.) and place names (“New York”, “South Korea” etc.).

library(quanteda)

Tokenize

toks <- tokens(data_corpus_inaugural)

Define multi-word expressions

Functions for tokens objects take a character vector, a dictionary or collocations as pattern. All those three can be used for multi-word expressions, but you have to be aware their differences.

Character vector

The most basic way to define multi-word expressions is separating words by whitespaces and wrap the character vector by phrase().

multiword <- c("United States", "New York")

Keyword-in-context

kwic() is useful to find multi-word expressions in tokens. If you are not sure if “United” and “States” are separated, check their positions (e.g. “434:435”).

head(kwic(toks, pattern = phrase(multiword)))
##                                                              
##  [1789-Washington, 434:435]            of the people of the |
##  [1789-Washington, 530:531]          more than those of the |
##       [1797-Adams, 525:526]     saw the Constitution of the |
##     [1797-Adams, 1717:1718]      to the Constitution of the |
##     [1797-Adams, 2481:2482] support the Constitution of the |
##   [1805-Jefferson, 441:442]       sees a taxgatherer of the |
##                                                       
##  United States | a Government instituted by themselves
##  United States | . Every step by which                
##  United States | in a foreign country.                
##  United States | , and a conscientious determination  
##  United States | , I entertain no doubt               
##  United States | ? These contributions enable us

Select tokens

Similarly, you can select or remove multi-word expression using tokens_select().

head(tokens_select(toks, pattern = phrase(multiword)))
## tokens from 6 documents.
## 1789-Washington :
## [1] "United" "States" "United" "States"
## 
## 1793-Washington :
## character(0)
## 
## 1797-Adams :
## [1] "United" "States" "United" "States" "United" "States"
## 
## 1801-Jefferson :
## character(0)
## 
## 1805-Jefferson :
## [1] "United" "States"
## 
## 1809-Madison :
## [1] "United" "States" "United" "States"

Compound tokens

tokens_compound() joins elements of multi-word expressions by underscore, so they become “United_States” and “New_York”.

comp_toks <- tokens_compound(toks, pattern = phrase(multiword))
head(tokens_select(comp_toks, pattern = c("United_States", "New_York")))
## tokens from 6 documents.
## 1789-Washington :
## [1] "United_States" "United_States"
## 
## 1793-Washington :
## character(0)
## 
## 1797-Adams :
## [1] "United_States" "United_States" "United_States"
## 
## 1801-Jefferson :
## character(0)
## 
## 1805-Jefferson :
## [1] "United_States"
## 
## 1809-Madison :
## [1] "United_States" "United_States"

Dictionary

Elements of multi-word expressions should be separately by whitespaces in a dictionary, but you do not use phrase() here.

Lookup dictionary

head(tokens_lookup(toks, dictionary = multiword_dict))
## tokens from 6 documents.
## 1789-Washington :
## [1] "country" "country"
## 
## 1793-Washington :
## character(0)
## 
## 1797-Adams :
## [1] "country" "country" "country"
## 
## 1801-Jefferson :
## character(0)
## 
## 1805-Jefferson :
## [1] "country"
## 
## 1809-Madison :
## [1] "country" "country"

Collocations

With textstat_collocations(), it is possible to discover multi-word expressions through statistical scoring of the associations of adjacent words.

Discover collocations

If textstat_collocations() is applied to a tokens object comprised only of capitalize words, it usually returns multi-word proper names.

col <- toks %>% 
       tokens_remove(stopwords("en")) %>% 
       tokens_select(pattern = "^[A-Z]", valuetype = "regex", 
                     case_insensitive = FALSE, padding = TRUE) %>% 
       textstat_collocations(min_count = 5, tolower = FALSE)
head(col)
##           collocation count count_nested length   lambda        z
## 1       United States   157            0      2 8.648770 28.48952
## 2  Federal Government    32            0      2 5.579973 21.89541
## 3        Almighty God    15            0      2 7.109404 18.22275
## 4       Chief Justice    13            0      2 8.846522 18.12255
## 5 Constitution United    19            0      2 4.037256 16.01611
## 6         North South     8            0      2 8.155828 15.41992

Compound collocations

Collocations are automatically recognized as multi-word expressions by tokens_compound() in case-sensitive fixed pattern matching. This is the fastest way to compound large numbers of multi-word expressions, but make sure that tolower = FALSE in textstat_collocations() to do this.

comp_toks2 <- tokens_compound(toks, pattern = col)
head(kwic(comp_toks2, pattern = c("United_States", "New_York")))
##                                                                          
##  [1789-Washington, 434]            of the people of the | United_States |
##  [1789-Washington, 529]          more than those of the | United_States |
##       [1797-Adams, 525]     saw the Constitution of the | United_States |
##      [1797-Adams, 1716]      to the Constitution of the | United_States |
##      [1797-Adams, 2479] support the Constitution of the | United_States |
##   [1805-Jefferson, 441]       sees a taxgatherer of the | United_States |
##                                       
##  a Government instituted by themselves
##  . Every step by which                
##  in a foreign country.                
##  , and a conscientious determination  
##  , I entertain no doubt               
##  ? These contributions enable us

You can use phrase() on collocations if more flexibility is needed. This is usually the case if you compound tokens from different corpus.

comp_toks3 <- tokens_compound(toks, pattern = phrase(col$collocation))
head(kwic(comp_toks3, pattern = c("United_States", "New_York")))
##                                                                          
##  [1789-Washington, 434]            of the people of the | United_States |
##  [1789-Washington, 529]          more than those of the | United_States |
##       [1797-Adams, 525]     saw the Constitution of the | United_States |
##      [1797-Adams, 1716]      to the Constitution of the | United_States |
##      [1797-Adams, 2479] support the Constitution of the | United_States |
##   [1805-Jefferson, 441]       sees a taxgatherer of the | United_States |
##                                       
##  a Government instituted by themselves
##  . Every step by which                
##  in a foreign country.                
##  , and a conscientious determination  
##  , I entertain no doubt               
##  ? These contributions enable us