Language Analyis Overview

Here we implement the application of text analyses to reveal association between stakeholders and Open Research (UNESCO) and business language.

This produces Fig 2c.

The code below includes the use of our control text corpus (book).

We use only words that are unique for each of the dictionaries. This enables us see the association and the divergence across the documents. We plotted the percentage of words that documents in the stakeholders groups share with the created by us dictionaries.

RATIO = number of words from one document present in one dictionary / total number of words in the document

Note: duplicates were not removed (it matters how many times a certain word occurs in a document)

Meaning of the columns in the final dataframe used for plotting:

• “document” - name of the document
• “present_BUS” - a word is present (1) or absent (0) in the business dictionary
• “present_UNESCO” - a word is present (1) or absent (0) in the UNESCO Recommendations
• “present_book” - a word is present (1) or absent (0) in the book
• “doc_pres” - total number of words present in the document (1)
• “sum” - a total number of words that are present in our dictionaries
• “stakeholder” - name of the stakeholder (funder/publisher/advocate/society/repository/journal)
• “proc_BUS” - % of number of words present in the certain document and business dictionary / total number of words in the certain document
• “proc_UNESCO” - % of number of words present in the certain document and UNESCO dictionary / total number of words in the certain document
• “proc_book” - % of number of words present in the certain document and book / total number of words in the certain document

R Setup and Packages

# Clearing R
rm(list=ls())


# Libraries used for text/data analysis
library(tidyverse) 
library(dplyr)
library(tidytext)

# Libraries used to create plots
library(ggplot2)

# Library to create a table when converting to html
library("kableExtra") 

Importing data

# Data: words, stakeholder, documents...
df_corpuses <- read_csv("./output/created_datasets/cleaned_data.csv")

Rows: 22822 Columns: 10
── Column specification ────────────────────────────────────────────────────────
Delimiter: ","
chr (10): txt, filename, name, doc_type, stakeholder, sentence_doc, orig_wor...

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

data_words <- df_corpuses

# Dictionary data 100 words per dictionary
# all_dict <- read_csv("./output/created_datasets/freq_dict_total_100.csv")
# All words
all_dict <- read_csv("./output/created_datasets/freq_dict_total_all.csv")

Rows: 9587 Columns: 6
── Column specification ────────────────────────────────────────────────────────
Delimiter: ","
chr (1): word
dbl (5): present_in_dict, present_BUS, present_UNESCO, present_book, sum

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

# Data preparation

# Changing name of the dictionary variable
new_dictionary <- all_dict
head(new_dictionary,3)

# A tibble: 3 × 6
  present_in_dict word     present_BUS present_UNESCO present_book   sum
            <dbl> <chr>          <dbl>          <dbl>        <dbl> <dbl>
1               1 deer               0              0            1     1
2               1 preserve           0              0            1     1
3               1 duck               0              0            1     1

# new_dictionary <- dictionaries
## Getting a data set with the words
data_words <- df_corpuses

# Merging
data_words <- data_words %>% 
  rename(document = name) %>% 
  select(document, stakeholder, word) %>% 
  left_join(new_dictionary, by = c("word" = "word")) # merging

# # HERE! I decided to remove words that did not appear in any of the dictionaries (20.09.2022)
# data_words <- data_words %>%
#   na.omit()


# Adding a column with stakeholder and word together to allow merging later
data_words$stake_word <- paste(data_words$stakeholder, "_", data_words$word)
# Adding a column with a document name and word together to allow merging later
data_words$doc_word <- paste(data_words$document, "_", data_words$word)
# Adding a column that will be used later to calculate the total of unique words in the document (dictionaries without removed words)
data_words$doc_pres <- 1

# Adding two columns that will be used later to calculate the total of unique words in the document (new dictionaries with removed common words)

# Replace NAs with 0 in all absence/presence columns
data_words_ND <- data_words %>%
  mutate_at(vars(present_in_dict, present_BUS, present_UNESCO, present_book, sum), ~replace_na(., 0)) # replacing NAs

# Select columns of my interest (stakeholders) and aggregate 

data_words_ND <- data_words_ND %>% 
  select(document, stakeholder, word, present_BUS, present_UNESCO, present_book, doc_pres, sum) # selecting columns, sum - column with the information about in how many dictionaries a certain word occurs

df_sum_pres_ND <- aggregate(x = data_words_ND[,4:8], by = list(data_words_ND$document), FUN = sum, na.rm = TRUE)
  
# head(df_sum_pres_ND, 3)

# By doing aggregate I lost info about the stakeholder the doc come from, I want to add it
df_doc_ord <- data_words_ND %>% 
  select(document, stakeholder) %>% 
  distinct(document, .keep_all = TRUE)
  
df_sum_pres_ND <- df_sum_pres_ND %>% 
  left_join(df_doc_ord, by = c("Group.1" = "document")) %>% 
  rename(document = Group.1)

# Creating % columns in a new df_sum_proc_ND data frame 
df_sum_proc_ND <- df_sum_pres_ND

df_sum_proc_ND$proc_BUS <-df_sum_proc_ND$present_BUS/df_sum_proc_ND$doc_pres*100

df_sum_proc_ND$proc_UNESCO <- df_sum_proc_ND$present_UNESCO/df_sum_proc_ND$doc_pres*100

df_sum_proc_ND$proc_book <- df_sum_proc_ND$present_book/df_sum_proc_ND$doc_pres*100


# Additional information about the data
# Stakeholders (2 from each of the stakeholders) that shared the highest no of words with UNESCO recommendation

UNESCO_stak_top <- df_sum_proc_ND %>% 
  group_by(stakeholder) %>% 
  arrange(desc(proc_UNESCO)) %>% 
  slice_head(n=2) %>% 
  select(stakeholder, document, proc_UNESCO)

UNESCO_stak_top %>% 
  kbl(caption = "Stakeholders that shared the higest no of words with UNESCO recommendation:") %>% 
  kable_classic("hover", full_width = T)

# Stakeholders (2 from each of the stakeholders) that shared the highest no of words with business dictionary

business_stak_top <- df_sum_proc_ND %>% 
  group_by(stakeholder) %>% 
  arrange(desc(proc_BUS)) %>% 
  slice_head(n=2) %>% 
  select(stakeholder, document, proc_BUS)

business_stak_top %>% 
  kbl(caption = "Stakeholders that shared the higest no of words with business dictionary:") %>% 
  kable_classic("hover", full_width = T)

# Stakeholders (2 from each of the stakeholders) that shared the highest no of words with book dictionary (control)

book_stak_top <-  df_sum_proc_ND %>% 
  group_by(stakeholder) %>% 
  arrange(desc(proc_book)) %>% 
  slice_head(n=2) %>% 
  select(stakeholder, document, proc_book)

book_stak_top %>% 
  kbl(caption = "Stakeholders that shared the higest no of words with book dictionary (control):") %>% 
  kable_classic("hover", full_width = T)

Graph generation

no <- nrow(df_sum_proc_ND)
no

[1] 129

# Plotting them separately book
df_sum_proc_ND_book = data.frame(
  document = rep(df_sum_proc_ND$document,1),
  stakeholder = rep(df_sum_proc_ND$stakeholder,1),
  type = c(rep("Book",no)),
  perc = c(df_sum_proc_ND$proc_book),
  perc2 = c(df_sum_proc_ND$proc_book))

sum_df_sum_proc_ND_book = 
  df_sum_proc_ND_book %>%
  group_by(stakeholder, type) %>%
  # dplyr::summarise(perc = mean(perc), SD = sd(perc2))
  dplyr::summarise(perc = median(perc), SD = sd(perc2))

`summarise()` has grouped output by 'stakeholder'. You can override using the
`.groups` argument.

fig_book <- ggplot() +
  geom_point(data = df_sum_proc_ND_book, aes(x = perc, y = stakeholder), alpha = 0.1, position = position_jitter()) +
  geom_pointrange(data = sum_df_sum_proc_ND_book, aes(x = perc, xmin = perc - SD, xmax = perc + SD, y = stakeholder)) +
  facet_grid(type~.) +
  labs(x = "Document words occuring in dictionary (%)", y = "") +
  scale_colour_discrete(guide = F) +
  theme_classic()
fig_book

Warning: It is deprecated to specify `guide = FALSE` to remove a guide. Please
use `guide = "none"` instead.

Past versions of unnamed-chunk-4-1.png

# Saving the figure
figure_name <- paste0("./output/Other_figures/language_book.png")
ggsave(filename = figure_name, fig_book + theme_bw(base_size = 5),
     width = 10, height = 5, dpi = 600, units = "in", device='png')  

Warning: It is deprecated to specify `guide = FALSE` to remove a guide. Please
use `guide = "none"` instead.

# Plotting them separately Business
df_sum_proc_ND_BUS = data.frame(
  document = rep(df_sum_proc_ND$document,1),
  stakeholder = rep(df_sum_proc_ND$stakeholder,1),
  type = c(rep("Business",no)),
  perc = c(df_sum_proc_ND$proc_BUS),
  perc2 = c(df_sum_proc_ND$proc_BUS))

sum_df_sum_proc_ND_BUS = 
  df_sum_proc_ND_BUS %>%
  group_by(stakeholder, type) %>%
  # dplyr::summarise(perc = mean(perc), SD = sd(perc2))
  dplyr::summarise(perc = median(perc), SD = sd(perc2))

`summarise()` has grouped output by 'stakeholder'. You can override using the
`.groups` argument.

fig_bus <- ggplot() +
  geom_point(data = df_sum_proc_ND_BUS, aes(x = perc, y = stakeholder), alpha = 0.1, position = position_jitter()) +
  geom_pointrange(data = sum_df_sum_proc_ND_BUS, aes(x = perc, xmin = perc - SD, xmax = perc + SD, y = stakeholder)) +
  facet_grid(type~.) +
  labs(x = "Document words occuring in dictionary (%)", y = "") +
  scale_colour_discrete(guide = F) +
  theme_classic()

fig_bus

Warning: It is deprecated to specify `guide = FALSE` to remove a guide. Please
use `guide = "none"` instead.

Past versions of unnamed-chunk-4-2.png

# Saving the figure
figure_name <- paste0("./output/Figure_2C/language_business.png")
ggsave(filename = figure_name, fig_bus + theme_bw(base_size = 5),
     width = 10, height = 5, dpi = 600, units = "in", device='png')  

Warning: It is deprecated to specify `guide = FALSE` to remove a guide. Please
use `guide = "none"` instead.

# Plotting them separately UNESCO
df_sum_proc_ND_UNESCO = data.frame(
  document = rep(df_sum_proc_ND$document,1),
  stakeholder = rep(df_sum_proc_ND$stakeholder,1),
  type = c(rep("UNESCO",no)),
  perc = c(df_sum_proc_ND$proc_UNESCO),
  perc2 = c(df_sum_proc_ND$proc_UNESCO))

sum_df_sum_proc_ND_UNESCO = 
  df_sum_proc_ND_UNESCO %>%
  group_by(stakeholder, type) %>%
  # dplyr::summarise(perc = mean(perc), SD = sd(perc2))
  dplyr::summarise(perc = median(perc), SD = sd(perc2))

`summarise()` has grouped output by 'stakeholder'. You can override using the
`.groups` argument.

fig_unesco <- ggplot() +
  geom_point(data = df_sum_proc_ND_UNESCO, aes(x = perc, y = stakeholder), alpha = 0.1, position = position_jitter()) +
  geom_pointrange(data = sum_df_sum_proc_ND_UNESCO, aes(x = perc, xmin = perc - SD, xmax = perc + SD, y = stakeholder)) +
  facet_grid(type~.) +
  labs(x = "Document words occuring in dictionary (%)", y = "") +
  scale_colour_discrete(guide = F) +
  theme_classic()
fig_unesco

Warning: It is deprecated to specify `guide = FALSE` to remove a guide. Please
use `guide = "none"` instead.

Past versions of unnamed-chunk-4-3.png

# Saving the figure
# UNCOMMENT TO SAVE FIGURE
# figure_name <- paste0("./output/Figure_2C/language_unesco.png")
# ggsave(filename = figure_name, fig_unesco  + theme_bw(base_size = 5),
#      width = 10, height = 5, dpi = 600, units = "in", device='png')  

Session information

sessionInfo()

R version 4.2.1 (2022-06-23)
Platform: x86_64-apple-darwin17.0 (64-bit)
Running under: macOS Big Sur ... 10.16

Matrix products: default
BLAS:   /Library/Frameworks/R.framework/Versions/4.2/Resources/lib/libRblas.0.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/4.2/Resources/lib/libRlapack.dylib

locale:
[1] en_GB.UTF-8/en_GB.UTF-8/en_GB.UTF-8/C/en_GB.UTF-8/en_GB.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] kableExtra_1.3.4 tidytext_0.3.4   forcats_0.5.2    stringr_1.4.1   
 [5] dplyr_1.0.10     purrr_0.3.5      readr_2.1.3      tidyr_1.2.1     
 [9] tibble_3.1.8     ggplot2_3.3.6    tidyverse_1.3.2  workflowr_1.7.0 

loaded via a namespace (and not attached):
 [1] fs_1.5.2            lubridate_1.8.0     bit64_4.0.5        
 [4] webshot_0.5.4       httr_1.4.4          rprojroot_2.0.3    
 [7] SnowballC_0.7.0     tools_4.2.1         backports_1.4.1    
[10] bslib_0.4.0         utf8_1.2.2          R6_2.5.1           
[13] DBI_1.1.3           colorspace_2.0-3    withr_2.5.0        
[16] tidyselect_1.2.0    processx_3.7.0      bit_4.0.4          
[19] compiler_4.2.1      git2r_0.30.1        textshaping_0.3.6  
[22] cli_3.4.1           rvest_1.0.3         xml2_1.3.3         
[25] labeling_0.4.2      sass_0.4.2          scales_1.2.1       
[28] callr_3.7.2         systemfonts_1.0.4   digest_0.6.29      
[31] rmarkdown_2.16      svglite_2.1.0       pkgconfig_2.0.3    
[34] htmltools_0.5.3     highr_0.9           dbplyr_2.2.1       
[37] fastmap_1.1.0       rlang_1.0.6         readxl_1.4.1       
[40] rstudioapi_0.14     farver_2.1.1        jquerylib_0.1.4    
[43] generics_0.1.3      jsonlite_1.8.3      vroom_1.6.0        
[46] tokenizers_0.2.3    googlesheets4_1.0.1 magrittr_2.0.3     
[49] Matrix_1.4-1        Rcpp_1.0.9          munsell_0.5.0      
[52] fansi_1.0.3         lifecycle_1.0.3     stringi_1.7.8      
[55] whisker_0.4         yaml_2.3.6          grid_4.2.1         
[58] parallel_4.2.1      promises_1.2.0.1    crayon_1.5.2       
[61] lattice_0.20-45     haven_2.5.1         hms_1.1.2          
[64] knitr_1.40          ps_1.7.1            pillar_1.8.1       
[67] reprex_2.0.2        glue_1.6.2          evaluate_0.16      
[70] getPass_0.2-2       modelr_0.1.9        vctrs_0.5.0        
[73] tzdb_0.3.0          httpuv_1.6.6        cellranger_1.1.0   
[76] gtable_0.3.1        assertthat_0.2.1    cachem_1.0.6       
[79] xfun_0.33           broom_1.0.1         janeaustenr_1.0.0  
[82] later_1.3.0         ragg_1.2.4          googledrive_2.0.0  
[85] viridisLite_0.4.1   gargle_1.2.1        ellipsis_0.3.2     

Session information

sessionInfo()

R version 4.2.1 (2022-06-23)
Platform: x86_64-apple-darwin17.0 (64-bit)
Running under: macOS Big Sur ... 10.16

Matrix products: default
BLAS:   /Library/Frameworks/R.framework/Versions/4.2/Resources/lib/libRblas.0.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/4.2/Resources/lib/libRlapack.dylib

locale:
[1] en_GB.UTF-8/en_GB.UTF-8/en_GB.UTF-8/C/en_GB.UTF-8/en_GB.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] kableExtra_1.3.4 tidytext_0.3.4   forcats_0.5.2    stringr_1.4.1   
 [5] dplyr_1.0.10     purrr_0.3.5      readr_2.1.3      tidyr_1.2.1     
 [9] tibble_3.1.8     ggplot2_3.3.6    tidyverse_1.3.2  workflowr_1.7.0 

loaded via a namespace (and not attached):
 [1] fs_1.5.2            lubridate_1.8.0     bit64_4.0.5        
 [4] webshot_0.5.4       httr_1.4.4          rprojroot_2.0.3    
 [7] SnowballC_0.7.0     tools_4.2.1         backports_1.4.1    
[10] bslib_0.4.0         utf8_1.2.2          R6_2.5.1           
[13] DBI_1.1.3           colorspace_2.0-3    withr_2.5.0        
[16] tidyselect_1.2.0    processx_3.7.0      bit_4.0.4          
[19] compiler_4.2.1      git2r_0.30.1        textshaping_0.3.6  
[22] cli_3.4.1           rvest_1.0.3         xml2_1.3.3         
[25] labeling_0.4.2      sass_0.4.2          scales_1.2.1       
[28] callr_3.7.2         systemfonts_1.0.4   digest_0.6.29      
[31] rmarkdown_2.16      svglite_2.1.0       pkgconfig_2.0.3    
[34] htmltools_0.5.3     highr_0.9           dbplyr_2.2.1       
[37] fastmap_1.1.0       rlang_1.0.6         readxl_1.4.1       
[40] rstudioapi_0.14     farver_2.1.1        jquerylib_0.1.4    
[43] generics_0.1.3      jsonlite_1.8.3      vroom_1.6.0        
[46] tokenizers_0.2.3    googlesheets4_1.0.1 magrittr_2.0.3     
[49] Matrix_1.4-1        Rcpp_1.0.9          munsell_0.5.0      
[52] fansi_1.0.3         lifecycle_1.0.3     stringi_1.7.8      
[55] whisker_0.4         yaml_2.3.6          grid_4.2.1         
[58] parallel_4.2.1      promises_1.2.0.1    crayon_1.5.2       
[61] lattice_0.20-45     haven_2.5.1         hms_1.1.2          
[64] knitr_1.40          ps_1.7.1            pillar_1.8.1       
[67] reprex_2.0.2        glue_1.6.2          evaluate_0.16      
[70] getPass_0.2-2       modelr_0.1.9        vctrs_0.5.0        
[73] tzdb_0.3.0          httpuv_1.6.6        cellranger_1.1.0   
[76] gtable_0.3.1        assertthat_0.2.1    cachem_1.0.6       
[79] xfun_0.33           broom_1.0.1         janeaustenr_1.0.0  
[82] later_1.3.0         ragg_1.2.4          googledrive_2.0.0  
[85] viridisLite_0.4.1   gargle_1.2.1        ellipsis_0.3.2