-
Notifications
You must be signed in to change notification settings - Fork 1
/
3_correlated_topic_model.R
310 lines (255 loc) · 10.8 KB
/
3_correlated_topic_model.R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
# Topic Modelling of CE Abstracts
# load data (optional)
MCE <- readRDS("data/MCE.RDS")
# Data preparation -------------------------------------------------------------------------
# load libraries
library(dplyr)
library(tibble)
# check for NAs
sum(is.na(MCE$AB))
# remove NAs
MCE <- MCE[!is.na(MCE$AB),]
doc_desc <- tibble(id = MCE$AU,
desc = MCE$AB)
# delete stopwords
library(tidytext)
doc_desc <- doc_desc %>%
unnest_tokens(word, desc) %>%
anti_join(stop_words)
# descriptions
worfreq <- doc_desc %>%
count(word, sort = TRUE)
# customize stopwords
my_stopwords <- tibble(word = c(as.character(1:10),"paper","study","research","based","results","result","elsevier","main","article","provide","due",
"role","rights","reserved", "main","proposed", "time", "data", "high", "literature", "article", "work", "order", "well", "authors",
"abstract", "current", "will", "focus", "provide", "approach", "considered", "based", "research", "rights", "reserved",
"2018","2016", "2017", "2015", "2014", "2013", "2012","2011","2010","2009","2008","2007", "2006","2005","2004","2003","2002",
"resulting", "resulted", "findings", "finding"))
# remove numbers
doc_desc <-doc_desc[-grep("\\b\\d+\\b", doc_desc$word),]
doc_desc <- doc_desc %>%
anti_join(my_stopwords)
word_counts <- doc_desc %>%
anti_join(my_stop_words) %>%
count(id, word, sort = TRUE) %>%
ungroup()
word_counts
# check co-ocurrences and correlations
library(widyr)
desc_word_pairs <- doc_desc %>%
pairwise_count(word, id, sort = TRUE, upper = FALSE)
desc_word_pairs
# create the document-term matrix
desc_dtm <- word_counts %>%
cast_dtm(id, word, n)
# Determining number of topics --------------------------------------------------
# Needs to be ACTIVATED as it takes several hours to fit the CTM
# # load library
# library(ldatuning)
#
# # run lda tuner
# result <- FindTopicsNumber(
# desc_dtm,
# topics = seq(from = 5, to = 100, by = 5),
# metrics = c("Griffiths2004", "CaoJuan2009", "Arun2010", "Deveaud2014"),
# method = "Gibbs",
# control = list(seed = 77),
# mc.cores = 2L,
# verbose = TRUE
# )
#
# # plot the result
# FindTopicsNumber_plot(result)
# Running the Correlated Topic Model -------------------------------------------------
# Needs to be ACTIVATED as it takes several hours to fit the CTM
# library(topicmodels)
#
# # be aware that running this model is time intensive
# ctm20 <- CTM(desc_dtm, k = 20, control = list(seed = 1234)) #
#
# # save result
# saveRDS(ctm20, file = "ctm20.RDS")
# Prepare data for plots -------------------------------------------------------
# load libraries
library(topicmodels)
library(dplyr)
library(tidytext)
library(purrr)
library(widyr)
library(tidyr)
library(ggplot2)
# load topic model
desc_ctm <- readRDS(file = "saves/ctm20_2000-2019.RDS")
# tidy the dataframe
tidy_ctm <- tidy(desc_ctm)
# define top terms
top_terms <- tidy_ctm %>%
group_by(topic) %>%
top_n(15, beta) %>%
ungroup() %>%
arrange(topic, -beta)
# Contribution plot ----------------------------------------------------------------------
# bring topics into the right order
top_terms <- as.data.frame(top_terms)
top_terms$topic_orderd <- ifelse(top_terms$topic == 7, 1,top_terms$topic ) # as we want to change more than one, we have to change the command a bit, so that we don't replace the new variables again with the old ones
top_terms$topic_orderd <- ifelse(top_terms$topic == 19, 2,top_terms$topic_orderd )
top_terms$topic_orderd <- ifelse(top_terms$topic == 2, 3,top_terms$topic_orderd )
top_terms$topic_orderd <- ifelse(top_terms$topic == 5, 4,top_terms$topic_orderd )
top_terms$topic_orderd <- ifelse(top_terms$topic == 17, 5,top_terms$topic_orderd )
top_terms$topic_orderd <- ifelse(top_terms$topic == 16, 6,top_terms$topic_orderd )
top_terms$topic_orderd <- ifelse(top_terms$topic == 9, 7,top_terms$topic_orderd )
top_terms$topic_orderd <- ifelse(top_terms$topic == 13, 8,top_terms$topic_orderd )
top_terms$topic_orderd <- ifelse(top_terms$topic == 10, 9,top_terms$topic_orderd )
top_terms$topic_orderd <- ifelse(top_terms$topic == 18, 10,top_terms$topic_orderd )
top_terms$topic_orderd <- ifelse(top_terms$topic == 1, 11,top_terms$topic_orderd )
top_terms$topic_orderd <- ifelse(top_terms$topic == 14, 12,top_terms$topic_orderd)
top_terms$topic_orderd <- ifelse(top_terms$topic == 3, 13,top_terms$topic_orderd )
top_terms$topic_orderd <- ifelse(top_terms$topic == 12, 14,top_terms$topic_orderd )
top_terms$topic_orderd <- ifelse(top_terms$topic == 4, 15,top_terms$topic_orderd )
top_terms$topic_orderd <- ifelse(top_terms$topic == 6, 16,top_terms$topic_orderd)
top_terms$topic_orderd <- ifelse(top_terms$topic == 15, 17,top_terms$topic_orderd )
top_terms$topic_orderd <- ifelse(top_terms$topic == 8, 18,top_terms$topic_orderd )
top_terms$topic_orderd <- ifelse(top_terms$topic == 11, 19,top_terms$topic_orderd )
top_terms$topic_orderd <- ifelse(top_terms$topic == 20, 20,top_terms$topic_orderd)
# assign correct order back to the original column
top_terms$topic <- top_terms$topic_orderd
top_terms %>%
mutate(term = reorder(term, beta)) %>%
group_by(topic, term) %>%
arrange(desc(beta)) %>%
ungroup() %>%
mutate(term = factor(paste(term, topic, sep = "__"),
levels = rev(paste(term, topic, sep = "__")))) %>%
ggplot(aes(term, beta, fill = as.factor(topic))) +
geom_col(show.legend = FALSE) +
coord_flip() +
scale_x_discrete(labels = function(x) gsub("__.+$", "", x)) +
labs(title = "Top 15 terms in each CTM topic",
x = NULL, y = expression(beta))+
facet_wrap(~ topic, ncol = 4, scales = "free")
# # DELETE in PAPER
# outdir <- "C:/Users/Sepp/Dropbox/CDLAB_Working Documents/"
# ggsave(file=paste(outdir, "Contribution.eps", sep=""),width = 15, height = 12, units = "in", dpi =600)
# Distribution plots --------------------------------------------------------------
# create ctm gamma
ctm_gamma <- tidy(desc_ctm, matrix = "gamma")
# Probability distribution all topics
ggplot(ctm_gamma, aes(gamma)) +
geom_histogram() +
scale_y_log10() +
labs(title = "Distribution of probabilities for all topics",
y = "Number of documents", x = expression(gamma))
# Probability distribution per topic
ggplot(ctm_gamma, aes(gamma, fill = as.factor(topic))) +
geom_histogram(show.legend = FALSE) +
facet_wrap(~ topic, ncol = 4) +
scale_y_log10() +
labs(title = "Distribution of probability for each topic",
y = "Number of documents", x = expression(gamma))
# Prevalence plot ------------------------------------------------------------------
# load libraries
library(stm)
library(furrr)
library(bigrquery)
library(tidyverse)
library(topicmodels)
library(dplyr)
library(tidytext)
library(purrr)
library(widyr)
library(tidyr)
library(knitr)
library(scales)
library(ggplot2)
#prepare tds
td_beta <- tidy(desc_ctm)
td_gamma <- tidy(topic_model, matrix = "gamma",
document_names = rownames(hacker_news_sparse))
# Let’s combine these to understand the topic prevalence
library(ggthemes)
top_terms <- td_beta %>%
arrange(beta) %>%
group_by(topic) %>%
top_n(5, beta) %>% # n can be varied to plot more or less terms
arrange(-beta) %>%
select(topic, term) %>%
summarise(terms = list(term)) %>%
mutate(terms = map(terms, paste, collapse = ", ")) %>%
unnest()
gamma_terms <- td_gamma %>%
group_by(topic) %>%
summarise(gamma = mean(gamma)) %>%
arrange(desc(gamma)) %>%
left_join(top_terms, by = "topic") %>%
mutate(topic = paste0("Topic ", topic),
topic = reorder(topic, gamma))
#rename gamma terms
gamma_terms <- as.data.frame(gamma_terms)
class(gamma_terms$topic)
gamma_terms$topic <- as.character(gamma_terms$topic)
gamma_terms[1,1] <- "Topic 1" # in order to be able to add names to the different topics
gamma_terms[2,1] <- "Topic 2"
gamma_terms[3,1] <- "Topic 3"
gamma_terms[4,1] <- "Topic 4"
gamma_terms[5,1] <- "Topic 5"
gamma_terms[6,1] <- "Topic 6"
gamma_terms[7,1] <- "Topic 7"
gamma_terms[8,1] <- "Topic 8"
gamma_terms[9,1] <- "Topic 9"
gamma_terms[10,1] <- "Topic 10"
gamma_terms[11,1] <- "Topic 11"
gamma_terms[12,1] <- "Topic 12"
gamma_terms[13,1] <- "Topic 13"
gamma_terms[14,1] <- "Topic 14"
gamma_terms[15,1] <- "Topic 15"
gamma_terms[16,1] <- "Topic 16"
gamma_terms[17,1] <- "Topic 17"
gamma_terms[18,1] <- "Topic 18"
gamma_terms[19,1] <- "Topic 19"
gamma_terms[20,1] <- "Topic 20"
gamma_terms$topic <- factor(gamma_terms$topic, levels = c("Topic 20",
"Topic 19",
"Topic 18",
"Topic 17",
"Topic 16",
"Topic 15",
"Topic 14",
"Topic 13",
"Topic 12",
"Topic 11",
"Topic 10",
"Topic 9",
"Topic 8",
"Topic 7",
"Topic 6",
"Topic 5",
"Topic 4",
"Topic 3",
"Topic 2",
"Topic 1"))
# plot
gamma_terms %>%
top_n(30, gamma) %>%
ggplot(aes(topic, gamma, label = terms, fill = topic)) +
geom_col(show.legend = FALSE) +
geom_text(hjust = 0, nudge_y = 0.0005, size = 4,
family = "IBMPlexSans") +
coord_flip() +
scale_y_continuous(expand = c(0,0),
limits = c(0, 0.25), #adjust the width of the plot with the percentage limit here
labels = percent_format()) +
theme_tufte(base_family = "IBMPlexSans", ticks = FALSE) +
theme(plot.title = element_text(size = 16,
family="IBMPlexSans-Bold"),
plot.subtitle = element_text(size = 13)) +
labs(x = NULL, y = expression(gamma),
title = "20 Topics by prevalence in Circular Economy research between 2000-2019",
subtitle = "With the top 5 words that contribute to each topic")
# show kable
gamma_terms %>%
select(topic, gamma, terms) %>%
kable(digits = 3,
col.names = c("Topic", "Expected topic proportion", "Top 7 terms"))
# create dataframe
CTM_top20 <- gamma_terms %>%
select(topic, gamma, terms)