IntroToTextAnalytics_Part3.R 4.64 KB
Newer Older
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
#
# Copyright 2017 Data Science Dojo
#    
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
# 
#    http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# 


#
# This R source code file corresponds to video 3 of the Data Science
# Dojo YouTube series "Introduction to Text Analytics with R" located 
# at the following URL:
#     https://www.youtube.com/watch?v=CQsyVDxK7_g     
#


# Install all required packages.
install.packages(c("ggplot2", "e1071", "caret", "quanteda", 
                   "irlba", "randomForest"))




# Load up the .CSV data and explore in RStudio.
spam.raw <- read.csv("spam.csv", stringsAsFactors = FALSE, fileEncoding = "UTF-16")
View(spam.raw)



# Clean up the data frame and view our handiwork.
spam.raw <- spam.raw[, 1:2]
names(spam.raw) <- c("Label", "Text")
View(spam.raw)



# Check data to see if there are missing values.
length(which(!complete.cases(spam.raw)))



# Convert our class label into a factor.
spam.raw$Label <- as.factor(spam.raw$Label)



# The first step, as always, is to explore the data.
# First, let's take a look at distibution of the class labels (i.e., ham vs. spam).
prop.table(table(spam.raw$Label))



# Next up, let's get a feel for the distribution of text lengths of the SMS 
# messages by adding a new feature for the length of each message.
spam.raw$TextLength <- nchar(spam.raw$Text)
summary(spam.raw$TextLength)



# Visualize distribution with ggplot2, adding segmentation for ham/spam.
library(ggplot2)

ggplot(spam.raw, aes(x = TextLength, fill = Label)) +
  theme_bw() +
  geom_histogram(binwidth = 5) +
  labs(y = "Text Count", x = "Length of Text",
       title = "Distribution of Text Lengths with Class Labels")



# At a minimum we need to split our data into a training set and a
# test set. In a true project we would want to use a three-way split 
# of training, validation, and test.
#
# As we know that our data has non-trivial class imbalance, we'll 
# use the mighty caret package to create a randomg train/test split 
# that ensures the correct ham/spam class label proportions (i.e., 
# we'll use caret for a random stratified split).
library(caret)
help(package = "caret")


# Use caret to create a 70%/30% stratified split. Set the random
# seed for reproducibility.
set.seed(32984)
indexes <- createDataPartition(spam.raw$Label, times = 1,
                               p = 0.7, list = FALSE)

train <- spam.raw[indexes,]
test <- spam.raw[-indexes,]


# Verify proportions.
prop.table(table(train$Label))
prop.table(table(test$Label))



# Text analytics requires a lot of data exploration, data pre-processing
# and data wrangling. Let's explore some examples.

# HTML-escaped ampersand character.
train$Text[21]


# HTML-escaped '<' and '>' characters. Also note that Mallika Sherawat
# is an actual person, but we will ignore the implications of this for
# this introductory tutorial.
train$Text[38]


# A URL.
train$Text[357]



# There are many packages in the R ecosystem for performing text
# analytics. One of the newer packages in quanteda. The quanteda
# package has many useful functions for quickly and easily working
# with text data.
library(quanteda)
help(package = "quanteda")


# Tokenize SMS text messages.
train.tokens <- tokens(train$Text, what = "word", 
                       remove_numbers = TRUE, remove_punct = TRUE,
                       remove_symbols = TRUE, remove_hyphens = TRUE)

# Take a look at a specific SMS message and see how it transforms.
train.tokens[[357]]


# Lower case the tokens.
train.tokens <- tokens_tolower(train.tokens)
train.tokens[[357]]


# Use quanteda's built-in stopword list for English.
# NOTE - You should always inspect stopword lists for applicability to
#        your problem/domain.
train.tokens <- tokens_select(train.tokens, stopwords(), 
                              selection = "remove")
train.tokens[[357]]


# Perform stemming on the tokens.
train.tokens <- tokens_wordstem(train.tokens, language = "english")
train.tokens[[357]]


# Create our first bag-of-words model.
train.tokens.dfm <- dfm(train.tokens, tolower = FALSE)


# Transform to a matrix and inspect.
train.tokens.matrix <- as.matrix(train.tokens.dfm)
View(train.tokens.matrix[1:20, 1:100])
dim(train.tokens.matrix)


# Investigate the effects of stemming.
colnames(train.tokens.matrix)[1:50]