Example Biterm Topic Model

Description

The object is a BTM topic model created with the BTM package. It was created on a subset of all CRAN packages, namely package which are part of the NaturalLanguageProcessing and MachineLearning task views.
Timepoint of creation was 2020-04-10.

Examples

library(BTM)
data(example_btm, package = 'textplot')
example_btm
str(example_btm)

Example word embedding matrix

Description

A matrix with 25-dimensional word embeddings, constructed upon the be_parliament_2020 dataset in the doc2vec R package

Examples

data(example_embedding, package = 'textplot')
head(example_embedding)

Example words emitted in a ETM text clustering model

Description

Example words emitted in a ETM text clustering model constructed upon the be_parliament_2020 dataset in the doc2vec R package

Examples

data(example_embedding_clusters, package = 'textplot')
head(example_embedding_clusters)
terminology <- split(example_embedding_clusters, example_embedding_clusters$cluster)
lapply(terminology, head, n = 5)

Example annotation of text using udpipe

Description

The object is a data.frame of the annotation of the text: "UDPipe provides tokenization, tagging, lemmatization and dependency parsing of raw text"

Examples

data(example_udpipe)
str(example_udpipe)

Plot function for a BTM object

Description

Plot biterms as a clustered graph. The graph is constructed by assigning each word to a topic and within a topic of words biterm frequencies are shown.

Usage

## S3 method for class 'BTM'
plot(
  x,
  biterms = terms(x, type = "biterms")$biterms,
  top_n = 7,
  which,
  labels = seq_len(x$K),
  title = "Biterm topic model",
  subtitle = list(),
  ...
)

Arguments

Value

an object of class ggplot

See Also

BTM, textplot_bitermclusters.default

Examples

library(igraph)
library(BTM)
library(ggraph)
library(ggforce)
library(concaveman)
data(example_btm, package = 'textplot')

model <- example_btm

plot(model, title = "BTM model", top_n = 3)
plot(model, title = "BTM model", top_n = 3, labels = 1:model$K)
plot(model, title = "BTM model", which = 7:15)
plot(model, title = "BTM model", subtitle = "First 5 topics",
     which = 1:5, top_n = 10)
plot(model, title = "Biterm topic model", subtitle = "First 8 topics",
     which = 1:8, top_n = 7)


topiclabels <- c("Garbage",
  "Data Mining", "Gradient descent", "API's",
  "Random Forests", "Stat models", "Text Mining / NLP",
  "GLM / GAM / Bayesian", "Machine learning", "Variable selection",
  "Regularisation techniques", "Optimisation", "Fuzzy logic",
  "Classification/Regression trees", "Text frequencies",
  "Neural / Deep learning", "Variable selection",
  "Text file handling", "Text matching", "Topic modelling")
plot(model, title = "Biterm topic model", subtitle = "some topics",
     top_n = 7,
     which = c(3, 4, 5, 6, 7, 9, 12, 16, 20),
     labels = topiclabels)


library(BTM)
library(data.table)
library(udpipe)
## Annotate text with parts of speech tags
data("brussels_reviews", package = "udpipe")
anno <- subset(brussels_reviews, language %in% "nl")
anno <- data.frame(doc_id = anno$id, text = anno$feedback, stringsAsFactors = FALSE)
anno <- udpipe(anno, "dutch", trace = 10)
## Get cooccurrences of nouns / adjectives and proper nouns
biterms <- as.data.table(anno)
biterms <- biterms[, cooccurrence(x = lemma,
                                  relevant = upos %in% c("NOUN", "PROPN", "ADJ"),
                                  skipgram = 2),
                     by = list(doc_id)]
## Build the BTM model
set.seed(123456)
x <- subset(anno, upos %in% c("NOUN", "PROPN", "ADJ"))
x <- x[, c("doc_id", "lemma")]
model <- BTM(x, k = 5, beta = 0.01, iter = 2000, background = TRUE,
             biterms = biterms, trace = 100)
plot(model)

Barplot of a frequency table using lattice

Description

Barplot of a frequency table using lattice

Usage

textplot_bar(x, ...)

## Default S3 method:
textplot_bar(
  x,
  panel = "Effect",
  total = sum(x),
  top = 40,
  col.panel = "lightgrey",
  col.line = "lightblue",
  lwd = 3,
  cextext = 0.5,
  addpct = FALSE,
  cexpct = 0.75,
  textpos = 3,
  pctpos = 1,
  v = NULL,
  col.abline = "red",
  ...
)

Arguments

Value

the result of a call to lattice::dotplot

Examples

data(brussels_listings, package = 'udpipe')
x <- table(brussels_listings$neighbourhood)
x <- sort(x)
textplot_bar(x,
 panel = "Locations", col.panel = "darkgrey", xlab = "Listings",
 cextext = 0.75, addpct = TRUE, cexpct = 0.5)


x <- sample(LETTERS, 1000, replace = TRUE)
textplot_bar(sort(table(x)), panel = "Frequencies", xlab = "Frequency",
   cextext = 0.75, main = "Freq stats")
textplot_bar(sort(table(x)), panel = "Frequencies", addpct = TRUE, top = 15)

## x can also be a data.frame where the first column
## is the label and the second column the frequency
x <- data.frame(l = LETTERS, amount = rnorm(26))
textplot_bar(x)
textplot_bar(x, v = 0)

Plot biterm cluster groups

Description

Plot biterms as a clustered graph. The graph is constructed by assigning each word to a topic and within a topic of words biterm frequencies are shown.

Usage

textplot_bitermclusters(x, ...)

## Default S3 method:
textplot_bitermclusters(
  x,
  biterms,
  which,
  labels = seq_len(length(table(biterms$topic))),
  title = "Biterm topic model",
  subtitle = list(),
  ...
)

Arguments

Value

an object of class ggplot

Examples

library(igraph)
library(ggraph)
library(concaveman)
library(ggplot2)
library(BTM)
data(example_btm, package = 'textplot')
group_terms   <- terms(example_btm, top_n = 3)
group_biterms <- example_btm$biterms$biterms


textplot_bitermclusters(x = group_terms, biterms = group_biterms)
textplot_bitermclusters(x = group_terms, biterms = group_biterms,
                        title = "BTM model", subtitle = "Topics 7-15",
                        which = 7:15, labels = seq_len(example_btm$K))

group_terms   <- terms(example_btm, top_n = 10)
textplot_bitermclusters(x = group_terms, biterms = group_biterms,
                        title = "BTM model", subtitle = "Topics 1-5",
                        which = 1:5, labels = seq_len(example_btm$K))

group_terms   <- terms(example_btm, top_n = 7)
topiclabels <- c("Garbage",
  "Data Mining", "Gradient descent", "API's",
  "Random Forests", "Stat models", "Text Mining / NLP",
  "GLM / GAM / Bayesian", "Machine learning", "Variable selection",
  "Regularisation techniques", "Optimisation", "Fuzzy logic",
  "Classification/Regression trees", "Text frequencies",
  "Neural / Deep learning", "Variable selection",
  "Text file handling", "Text matching", "Topic modelling")
textplot_bitermclusters(x = group_terms, biterms = group_biterms,
                        title = "Biterm topic model", subtitle = "some topics",
                        which = c(3, 4, 5, 6, 7, 9, 12, 16, 20),
                        labels = topiclabels)

Plot term cooccurrences as a network

Description

Plot term cooccurrences in a graph structure

Usage

textplot_cooccurrence(x, ...)

## Default S3 method:
textplot_cooccurrence(
  x,
  terms,
  top_n = 50,
  title = "Term cooccurrences",
  subtitle = list(),
  vertex_color = "darkgreen",
  edge_color = "grey",
  base_family = "",
  ...
)

Arguments

Value

an object of class ggplot

Examples

library(udpipe)
library(igraph)
library(ggraph)
library(ggplot2)
data(brussels_reviews_anno, package = 'udpipe')
x <- subset(brussels_reviews_anno, xpos %in% "JJ" & language %in% "fr")
x <- cooccurrence(x, group = "doc_id", term = "lemma")

textplot_cooccurrence(x, top_n = 25, subtitle = "showing only top 25")
textplot_cooccurrence(x, top_n = 25, title = "Adjectives",
                      vertex_color = "orange", edge_color = "black",
                      fontface = "bold")

Plot sparse term correlations as a graph structure

Description

Plot sparse term correlations as a graph structure. Uses the glasso procedure (glasso::glassopath) to reduce the correlation matrix to retain only the relevant correlations and next visualises these sparse correlations.

Usage

textplot_correlation_glasso(x, ...)

## Default S3 method:
textplot_correlation_glasso(
  x,
  n = 1000,
  exclude_zero = TRUE,
  label.cex = 1,
  node.width = 0.5,
  ...
)

Arguments

Value

an object of class ggplot

Examples

library(udpipe)
library(qgraph)
library(glasso)
data(brussels_reviews_anno, package = 'udpipe')
x <- subset(brussels_reviews_anno, xpos %in% "NN" & language %in% "fr" & !is.na(lemma))
x <- document_term_frequencies(x, document = "doc_id", term = "lemma")
dtm <- document_term_matrix(x)
dtm <- dtm_remove_lowfreq(dtm, maxterms = 60)

m <- dtm_cor(dtm)
textplot_correlation_glasso(m, exclude_zero = TRUE)

textplot_correlation_glasso(m, exclude_zero = FALSE)

Document/Term Correlation Plot

Description

Plots the highest occurring correlations among terms.
This is done by plotting the terms into nodes and the correlations between the terms as lines between the nodes. Lines of the edges are proportional to the correlation height. This uses the plot function for graphNEL objects (using the Rgraphviz package)

Usage

textplot_correlation_lines(x, ...)

## Default S3 method:
textplot_correlation_lines(
  x,
  terms = colnames(x),
  threshold = 0.05,
  top_n,
  attrs = textplot_correlation_lines_attrs(),
  terms_highlight,
  label = FALSE,
  cex.label = 1,
  col.highlight = "red",
  lwd = 1,
  ...
)

Arguments

Value

invisibly the plot

Examples

## Construct document/frequency/matrix
library(graph)
library(Rgraphviz)
library(udpipe)
data(brussels_reviews_anno, package = 'udpipe')
exclude <- c(32337682L, 27210436L, 26820445L, 37658826L, 33661134L, 48756422L,
  23454554L, 30461127L, 23292176L, 32850277L, 30566303L, 21595142L,
  20441279L, 38097066L, 28651065L, 29011387L, 37316020L, 22135291L,
  40169379L, 38627667L, 29470172L, 24071827L, 40478869L, 36825304L,
  21597085L, 21427658L, 7890178L, 32322472L, 39874379L, 32581310L,
  43865675L, 31586937L, 32454912L, 34861703L, 31403168L, 35997324L,
  29002317L, 33546304L, 47677695L)
dtm <- brussels_reviews_anno
dtm <- subset(dtm, !doc_id %in% exclude)
dtm <- subset(dtm, xpos %in% c("NN") & language == "nl" & !is.na(lemma))
dtm <- document_term_frequencies(dtm, document = "doc_id", term = "lemma")
dtm <- document_term_matrix(dtm)
dtm <- dtm_remove_lowfreq(dtm, minfreq = 5)
dtm <- dtm_remove_tfidf(dtm, top = 500)

## Plot top 20 correlations, having at least a correlation of 0.01
textplot_correlation_lines(dtm, top_n = 25, threshold = 0.01)

## Plot top 20 correlations
textplot_correlation_lines(dtm, top_n = 25, label = TRUE, lwd = 5)

## Plot top 20 correlations and highlight some terms
textplot_correlation_lines(dtm, top_n = 25, label = TRUE, lwd = 5,
                           terms_highlight = c("prijs", "privacy"),
                           main = "Top correlations in topic xyz")

## Plot top 20 correlations and highlight + increase some terms
textplot_correlation_lines(dtm, top_n = 25, label = TRUE, lwd=5,
                           terms_highlight = c(prijs = 0.8, privacy = 0.1),
                           col.highlight = "red")

## Plot correlations between specific terms
w <- dtm_colsums(dtm)
w <- head(sort(w, decreasing = TRUE), 100)
textplot_correlation_lines(dtm, terms = names(w), top_n = 20, label = TRUE)

attrs <- textplot_correlation_lines_attrs()
attrs$node$shape <- "rectangle"
attrs$edge$color <- "steelblue"
textplot_correlation_lines(dtm, top_n = 20, label = TRUE,
                           attrs = attrs)

Document/Term Correlation Plot graphical attributes

Description

Document/Term Correlation Plot graphical attributes

Usage

textplot_correlation_lines_attrs(fontsize = 25)

Arguments

Value

a list with graph visualisation elements used by textplot_correlation_lines

Examples

textplot_correlation_lines_attrs()

Plot output of a dependency parser

Description

Plot output of a dependency parser. This plot takes one sentence and shows for the sentence, the words, the parts of speech tag and the dependency relationship between the words.

Usage

textplot_dependencyparser(x, ...)

## Default S3 method:
textplot_dependencyparser(
  x,
  title = "Dependency Parser",
  subtitle = "tokenisation, parts of speech tagging & dependency relations",
  vertex_color = "darkgreen",
  edge_color = "red",
  size = 3,
  base_family = "",
  layout = "linear",
  ...
)

Arguments

Value

an object of class ggplot

See Also

udpipe

Examples

library(udpipe)
library(ggraph)
library(ggplot2)
library(igraph)

x <- udpipe("The economy is weak but the outlook is bright", "english")
textplot_dependencyparser(x)

x <- udpipe("His speech about marshmallows in New York is utter bullshit", "english")
textplot_dependencyparser(x, size = 4)

x <- udpipe("UDPipe provides tokenization, tagging, lemmatization and
             dependency parsing of raw text", "english")
textplot_dependencyparser(x, size = 4)


data("example_udpipe", package = "textplot")
textplot_dependencyparser(example_udpipe, size = 4)

Plot word embeddings in 2D

Description

This plot displays words in 2 dimensions, optionally grouped by cluster.
This allows to visualise embeddings which are reduced by dimensionality reduction techniques like UMAP, t-SNE, PCA or similar techniques. It allows to highlight the words by groups and is a good way to visualise a small sets of word or topic embeddings.

Usage

textplot_embedding_2d(x, ...)

## Default S3 method:
textplot_embedding_2d(
  x,
  title = "Embedding plot in 2D",
  subtitle = list(),
  encircle = FALSE,
  points = FALSE,
  alpha = 0.4,
  ...
)

Arguments

Value

an object of class ggplot

Examples

library(ggplot2)
library(ggrepel)
library(ggalt)
##
## Generate some fake embeddings
##   probably you want to use word2vec::word2vec(...) + uwot::umap(...)
embeddings <- matrix(runif(26 * 2), nrow = 26, ncol = 2, dimnames = list(letters))
x <- data.frame(term = rownames(embeddings), x = embeddings[, 1], y = embeddings[, 2])

## 2D plot
textplot_embedding_2d(x)

## 2D plot with groups
x$group <- sample(c("clustera", "clusterb", "clusterc"), size = 26, replace = TRUE)
textplot_embedding_2d(x)

## 2D plot with groups and weights for each word
x$weight <- runif(nrow(x))
textplot_embedding_2d(x)
textplot_embedding_2d(x, points = TRUE)

## 2D plot with groups and weights for each word and different types of points
x$type <- sample(c("word", "center"), size = 26, replace = TRUE)
x$type <- factor(x$type, levels = c("word", "center"))
textplot_embedding_2d(x, points = TRUE)
textplot_embedding_2d(x, title = "Embedding plot in 2D", subtitle = "example")

## Encircle the words belonging to each group
textplot_embedding_2d(x, title = "Embedding plot in 2D", subtitle = "example",
                      encircle = TRUE, alpha = 0.2)