Practical Natural Language Processing with

Python With Case Studies from Industries Using

Text Data at Scale 1st Edition Mathangi Sri pdf
download
https://textbookfull.com/product/practical-natural-language-
processing-with-python-with-case-studies-from-industries-using-
text-data-at-scale-1st-edition-mathangi-sri/

Download more ebook from https://textbookfull.com

We believe these products will be a great fit for you. Click
the link to download now, or visit textbookfull.com
to discover even more!

Natural Language Processing with Spark NLP Learning to

Understand Text at Scale 1st Edition Alex Thomas

https://textbookfull.com/product/natural-language-processing-
with-spark-nlp-learning-to-understand-text-at-scale-1st-edition-
alex-thomas/

Text Analytics with Python - A Practitioner's Guide to

Natural Language Processing 2nd Edition Dipanjan Sarkar

https://textbookfull.com/product/text-analytics-with-python-a-
practitioners-guide-to-natural-language-processing-2nd-edition-
dipanjan-sarkar/

Natural Language Processing with Python Cookbook 1st

Edition Krishna Bhavsar

https://textbookfull.com/product/natural-language-processing-
with-python-cookbook-1st-edition-krishna-bhavsar/

Natural language processing with TensorFlow Teach

language to machines using Python s deep learning
library 1st Edition Thushan Ganegedara

https://textbookfull.com/product/natural-language-processing-
with-tensorflow-teach-language-to-machines-using-python-s-deep-
learning-library-1st-edition-thushan-ganegedara/
Applied Text Analysis with Python Enabling Language
Aware Data Products with Machine Learning 1st Edition
Benjamin Bengfort

https://textbookfull.com/product/applied-text-analysis-with-
python-enabling-language-aware-data-products-with-machine-
learning-1st-edition-benjamin-bengfort/

Applied Natural Language Processing with Python:

Implementing Machine Learning and Deep Learning
Algorithms for Natural Language Processing 1st Edition
Taweh Beysolow Ii
https://textbookfull.com/product/applied-natural-language-
processing-with-python-implementing-machine-learning-and-deep-
learning-algorithms-for-natural-language-processing-1st-edition-
taweh-beysolow-ii/

Text Analytics with Python A Practical Real World

Approach to Gaining Actionable Insights from Your Data
1st Edition Dipanjan Sarkar

https://textbookfull.com/product/text-analytics-with-python-a-
practical-real-world-approach-to-gaining-actionable-insights-
from-your-data-1st-edition-dipanjan-sarkar/

Advanced Analytics with Spark Patterns for Learning

from Data at Scale 2nd Edition Sandy Ryza

https://textbookfull.com/product/advanced-analytics-with-spark-
patterns-for-learning-from-data-at-scale-2nd-edition-sandy-ryza/

Machine Learning with PySpark: With Natural Language

Processing and Recommender Systems 1st Edition Pramod
Singh

https://textbookfull.com/product/machine-learning-with-pyspark-
with-natural-language-processing-and-recommender-systems-1st-
edition-pramod-singh/
Mathangi Sri

Practical Natural Language Processing

with Python
With Case Studies from Industries Using Text Data
at Scale
1st ed.
Mathangi Sri
Bangalore, Karnataka, India

Any source code or other supplementary material referenced by the

author in this book is available to readers on GitHub via the book’s
product page, located at www.​apress.​com/​978-1-4842-6245-0. For
more detailed information, please visit www.​apress.​com/​source-code.

ISBN 978-1-4842-6245-0 e-ISBN 978-1-4842-6246-7

https://doi.org/10.1007/978-1-4842-6246-7

© Mathangi Sri 2021

Apress Standard

Trademarked names, logos, and images may appear in this book. Rather
than use a trademark symbol with every occurrence of a trademarked
name, logo, or image we use the names, logos, and images only in an
editorial fashion and to the benefit of the trademark owner, with no
intention of infringement of the trademark. The use in this publication
of trade names, trademarks, service marks, and similar terms, even if
they are not identified as such, is not to be taken as an expression of
opinion as to whether or not they are subject to proprietary rights.

The publisher, the authors and the editors are safe to assume that the
advice and information in this book are believed to be true and accurate
at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, express or implied, with respect to the material
contained herein or for any errors or omissions that may have been
made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

Distributed to the book trade worldwide by Springer Science+Business

Media New York, 1 New York Plaza, Suite 4600, New York, NY 10004-
1562, USA. Phone 1-800-SPRINGER, fax (201) 348-4505, e-mail orders-
[email protected], or visit www.springeronline.com. Apress Media,
LLC is a California LLC and the sole member (owner) is Springer
Science + Business Media Finance Inc (SSBM Finance Inc). SSBM
Finance Inc is a Delaware corporation.
To my loving daughter Sahana
who inspires me every day with her unending spirit to learn
Introduction
I am fortunate to have had the exposure and opportunity to solve
complex NLP problems that benefited various businesses across
geographies. This book comes from my learnings and hence it is a
practitioner view of solving text problems. Solving NLP problems
involves a certain combination of creativity and technical knowledge.
Sometimes the best deep learning methods do not solve a problem as
well as simple solutions do. I am always reminded of Occam’s razor,
which states that when there are alternatives available, the simplest
one possibly solves the problem best.
In my opinion, the answer to any problem lies in the data. This is
why the first chapter talks about text data. I cover different types of text
data and the information that can be extracted from this data. Chapters
2, 3, 4, and 5 focus on different industries and the common text mining
problems faced in these domains. As I cover the use cases in these
chapters, I also cover the corresponding text mining concepts along
with the relevant code and Python libraries . Chapter 2 takes a quick
view of the customer service industry with a deep dive into the
conversation corpus. It highlights the various types of information that
can be extracted from the conversation corpus along with detailed
code. I also cover bag of words, vectorization, the rules-based approach,
and supervised learning methods in great detail. Chapter 3 focuses on a
very popular text mining use case: review mining. I delve into
sentiment analysis and the various facets of opinion mining. I cover
sentiment analysis using unsupervised and supervised techniques. I
introduce neural networks for sentiment analysis in this chapter.
Chapter 4 provides an overview of techniques used in the banking and
financial services industries . Traditionally, banking uses structured
data to drive decisions. Of late, text use cases are being explored in this
arena and I talk about one in detail. I explore named entity recognizers
in detail in this chapter using unsupervised techniques and embedding
based neural networks. Chapter 5 is dedicated to virtual assistants . I
explore techniques to build bots using state-of-the-art neural network
architectures. This chapter also introduces natural language generation
concepts.
Acknowledgments
I want to thank my husband, Satish. I brainstormed with him several
times during the course of this book on the technical and business use
cases. I also want to thank the team at Apress for providing adequate
reviews and guidance on the content. I learned all the concepts on the
job, so I thank all the people with whom I have had the privilege to
work with and learn from.
Table of Contents
Chapter 1:​Types of Data
Search
Reviews
Social Media Posts/​Blogs
Chat Data
Personal Chats
Business Chats and Voice Call Data
SMS Data
Content Data
IVR Utterance Data
Useful Information from Data
Chapter 2:​NLP in Customer Service
Voice Calls
Chats
Tickets Data
Email data
Voice of Customer
Intent Mining
Top Words to Understand Intents
Word Cloud
Rules to Classify Topics
Supervised Learning Using Machine Learning
Getting Manually Labelled Data
Word Tokenization
Term-Document Matrix
 Data Normalization
Replacing Certain Patterns
Identifying Issue Lines
Top Customer Queries
Top CSAT Drivers
Top NPS Drivers
Insights into Sales Chats
Top Products for Sales Chats
Reasons for Non-Purchase
Survey Comments Analysis
Mining Voice Transcripts
Acoustic Model
Language Model
Chapter 3:​NLP in Online Reviews
Sentiment Analysis
Emotion Mining
Approach 1:​Lexicon-Based Approach
Approach 2:​Rules-Based Approach
Observation 1
Observation 2
Observation 3
Observation 4
Overall Score
Implementing the Observations
Sentiment Analysis Libraries
 Approach 3:​Machine-Learning Based Approach (Neural
Network)
Corpus Features
Things to Improve
Attribute Extraction
Step 1:​Using Regex to Normalize
Step 2:​Extracting Noun Forms
Step 3:​Creating a Mapping File
Step 4:​Mapping Each Review to an Attribute
Step 5:​Analyzing Brands
Chapter 4:​NLP in Banking, Financial Services, and Insurance
(BFSI)
NLP in Fraud
Method 1:​Using Existing Libraries
Method 2:​Extracting Noun Phrases
Method 3:​Training Your Own Model
Applying the Model
Other Use Cases of NLP in Banking, Financial Services, and
Insurance
SMS Data
Natural Language Generation in Banks
Chapter 5:​NLP in Virtual Assistants
Types of Bots
The Classic Approach
Quick Overview of LSTM
Applying LSTM
Time-Distributed Layer
 Approach 2 - The Generating Responses Approach
Encoder-Decoder Architecture
Dataset
Implementing the Architecture
Encoder-Decoder Training
Encoder Output
Decoder Input
Preprocessing
Bidirectional LSTM
BERT
Language Models and Fine-Tuning
Overview of BERT
Fine-Tuning BERT for a Classifier
Further Nuances in Building Conversational Bots
Single-Turn vs.​Multi-Turn Conversations
Multi-Lingual Bots
Index
About the Author
Mathangi Sri
is a renowned data science leader in
India. She has 11 patent grants and 20+
patents published in the area of intuitive
customer experience, indoor positioning,
and user profiles. She has 16+ years of a
proven track record in building world-
class data science solutions and
products. She is adept in machine
learning, text mining, and NLP
technologies and tools. She has built data
science teams across large organizations
like Citibank, HSBC, and GE as well as
tech startups like 247.​ai, PhonePe, and
Gojek. She advises startups, enterprises, and venture capitalists on data
science strategy and roadmaps. She is an active contributor on machine
learning to many premier institutes in India. She was recognized as one
of “The Phenomenal SHE” by the Indian National Bar Association in
2019.
About the Technical Reviewer
Manohar Swamynathan
is a data science practitioner and an avid
programmer, with over 14+ years of
experience in various data science-
related areas, including data
warehousing, business intelligence,
analytical tool development, ad-hoc
analysis, predictive modeling, data
science product development,
consulting, formulating strategy, and
executing analytics programming. His
career has covered the life cycle of data
across different domains such as US
mortgage banking, retail/e-commerce, insurance, and industrial IoT.
He’s also involved in the technical review of books about data science
using Python and R. He has a bachelor’s degree with a specialization in
physics, mathematics, and computers, and a Master’s degree in project
management. He’s currently living in Bengaluru, the silicon valley of
India.
© Mathangi Sri 2021
M. Sri, Practical Natural Language Processing with Python
https://doi.org/10.1007/978-1-4842-6246-7_1

1. Types of Data
Mathangi Sri1
(1) Bangalore, Karnataka, India

Natural language processing (NLP) is a field that helps humans

communicate with computers naturally. It is a shift from the era when
humans had to “learn” to use computers to computers being trained to
understand humans. It is a branch of artificial intelligence (AI) that
deals with language. The field dates back to the 1950s when a lot of
research was undertaken in the machine translation area. Alan Turing
predicted that by the early 2000s computers would be able to
flawlessly understand and respond in natural language that you won’t
be able to distinguish between humans and computers. We are far from
that benchmark in the field of NLP. However, some argue that this may
not even be the right lens to measure achievements in the field. Be that
as it may, NLP is central to the success of many businesses. It is very
difficult to imagine life without Google search, Alexa, YouTube
recommendations, and so on. NLP has become ubiquitous today.
In order to understand this branch of AI better, let’s start with the
fundamentals. Fundamental to any data science field is data. Hence
understanding text data and various forms of it is at the heart of
performing natural language processing. Let’s start with some of the
most familiar daily sources of text data, from the angle of commercial
usage:
Search
Reviews
Social media posts/blogs
Chat data (business-to-consumer and consumer-to-consumer)
 SMS data
Content data (news/videos/books)
IVR utterance data

Search
Search is one of the most widely used data sources from a customer
angle. All search engine searches, whether a universal search engine or
a search inside a website or an app, use at the core indexing, retrieval,
and relevance-ranking algorithms. Search, also referred to as a query, is
typically made up of short sentences of two or three words. Search
engine results are approximate and they don’t necessarily need to be
bang on with their results. For a query, multiple options are always
presented as results. This user interface transfers the onus of finding
the answer back to the user. Recount the number of times you have
modified your query because you were not satisfied with the result. It’s
unlikely that you blamed the performance of the engine. You focused
your attention on modifying your query.

Reviews
Reviews are possibly the most widely analyzed data. Since this data is
available openly or is easy to extract with web crawling, many
organizations use this data. Reviews are very free flowing in nature and
are very unstructured. Review mining is core to e-commerce companies
like Amazon, Flipkart, eBay, and so on. Review sites like IMDB and
Tripadvisor also have reviews data at their core. There are other
organizations/vendors that provide insights on reviews collected by
these companies. Figure 1-1 shows sample review data from
www.amazon.in/dp/B0792KTHKK/ref=gw-hero-PC-dot-
news-sketch?pf_rd_p=865a7afb-79a5-499b-82de-
731a580ea265&pf_rd_r=TGGMS83TD4VZW7KQQBF3.
Figure 1-1 Sample Amazon review
Note that the above review highlights the features that are
important to the user: the scope of the product (music), the search
efficiency, the speaker, and its sentiment. But we also get to know
something about the user, such as the apps they care about. We could
also profile the user on how objective or subjective they are.
As a quick, fun exercise, look at the long review from Amazon in
Figure 1-2 and list the information you can extract from the review in
the following categories: product features, sentiment, about the user,
user sentiment, and whether the user is a purchaser.
Figure 1-2 Extract some data from this review.

Social Media Posts/Blogs

Social media posts and blogs are widely researched, extracted, and
analyzed, like reviews. Tweets and other microblogs are short and
hence could seem easily extractable. However, tweets, depending on
use cases, can carry a lot of noise. From my experience, on average only
1 out of every 100 tweets contains useful information on a given
concept of interest. This is especially true in cases of analyzing
sentiments for brands using Twitter data. In this research paper on
sentiment analysis, only 20% of tweets in English and 10% of tweets in
Turkish were found to be useful after collecting tweets for the topic:
www.researchgate.net/profile/Serkan_Ayvaz/publicat
ion/320577176_Sentiment_Analysis_on_Twitter_A_Text
_Mining_Approach_to_the_Syrian_Refugee_Crisis/link
s/5ec83c79299bf1c09ad59fb4/Sentiment-Analysis-on-
Twitter-A-Text-Mining-Approach-to-the-Syrian-
Refugee-Crisis.pdf. Hence looking for the right tweet in a corpus
of tweets is a key to successfully mining Twitter or Facebook posts.
Let’s take an example from https://twitter.com/explore:

Night Santa Cruz boardwalk and ocean

Took me while to get settings right. .....

Camera: pixel 3

Setting: raw, 1...https://t.co/XJfDq4WCuu

@Google @madebygoogle could you guys hook me up

with the upcoming Pixel 4XL for my pixel IG. Just
trying to stay ah...https://t.co/LxBHIRkGG1

China's bustling cities and countryside were

perfect for a smartphone camera test. I pitted the
#HuaweiP30Pro again...https://t.co/Cm79GQJnBT

#sun #sunrise #morningsky #glow #rooftop

#silohuette madebygoogle google googlepixel #pixel
#pixel3 #pixel3photos...https://t.co/vbScNVPjfy

RT @kwekubour: With The Effortlessly Fine, @acynam

x Pixel 3

Get A #Google #Pixel3 For $299, #Pixel3XL For $399

With Activation In These Smoking Hot
#Dealshttps://t.co/ydbadB5lAn via @HotHardware

I purchased pixel 3 on January 26 2019 i started

facing call drops issue and it is increasing day
by day.i dont kn...https://t.co/1LTw9EdYzp
As you can see in this example, which displays sample tweets for Pixel
3, the content spans deals, reviews of the phone, amazing shots taken
from the phone, someone awaiting the Pixel 4, and so on. In fact, if you
want to understand the review or sentiment associated with Pixel 3,
only 1 out of the 8 tweets is relevant.
A microblog’s data can contain power-packed information about a
topic. In the above example of Pixel 3, you can find the following: the
most liked or disliked features, the influence of location on the topic,
the perception change over time, the impact of advertisements, the
perception of advertisements for Pixel 3, and what kind of users like or
dislike the product. Twitter can be mined as a “leading indicator” for
various events, such as if a stock price of a particular company can be
predicted if there is significant news about the company. The research
paper at
www.sciencedirect.com/science/article/pii/S2405918
817300247 describes how Twitter data was used to correlate the
movements of the FTSE Index (Financial Times Stock Exchange Index,
an index of the 100 most capitalized companies on the London Stock
Exchange) before, during, and after local elections in the United
Kingdom.

Chat Data
Personal Chats
Personal chats are the classic everyday corpus of WhatsApp chat or
Facebook or any other messenger service. They are definitely one of the
richest sources of information to understand user behavior, more in the
friends-and-family circle. They are filled with a lot of noise that needs
to be weeded out, like you saw with the Twitter data. Only a small
portion of the corpus is relevant for extracting useful commercial
information. The incidence rate of this commercially useful information
is not very high. That is to say, it has a low signal-to-noise ratio. The
paper at
www.aaai.org/ocs/index.php/ICWSM/ICWSM18/paper/vie
w/17865/17043 studied openly submitted data from various
WhatsApp chats. Figure 1-3 shows a word cloud of the WhatsApp
groups analyzed by the paper.

Figure 1-3 WhatsApp word cloud

Also the data privacy guidelines of messenger apps may not permit
mining personal chats for commercial purposes. Some of the personal
chats have functionality where a business can interact with the user,
and I will cover that as part of the next section

Business Chats and Voice Call Data

Business chats, also referred to as live chats, are conversations that
consumers have with a business on their website or app. Typically
users reach out to chat agents about the issue they face in using a
product or service. They may also discuss the product before making a
purchase decision. Business chats are fairly rich in information, more so
on the commercial preferences of the user. Look at the following
example of a chat:

system Thank you for choosing Best Telco. A representative will be with
you shortly.
system You are now chatting with ‘Max’
customer hi max
agent Thank you for contacting Best Telco. My name is Max.
agent Hello, I see that I’m chatting with Mrs. Sara and you’ve provided
XXXXXXX as the number associated with your account. Is that
correct?
customer yes. that is correct. I just received email saying my service price
was going up to $70/month but on your web site, it’s only $30.
What’s the deal with that?
agent Thank you for the information Mrs. Sara.
agent I’m sorry to hear about the increase, our billing department can
verify on the changes of the prices.
customer who do I need to talk to?
agent I will provide you the billing department number. Sorry
customer thank you
Agent The number for Billing is: XXXXXXXX, they are open Monday -
Friday: 8:00 a.m. - 6:00 p.m. ET. Saturday/Sunday/Holidays:
Closed
Customer ok thanks. good bye
A lot of information can be cleanly taken from the above corpus. The
user name, their problem, the fact that the user is responsive to emails,
the user is also sensitive to price, the user’s sentiment, the courtesy of
the agent, the outcome of the chat, the resolution provided by the
agent, and the different departments of Best Telco.
Also, note how the data is laid out: it’s free flow text from the
customer. But the chat agent plays a critical role in directing the chat.
The initial lines talk about the issue, then the agent presents a
resolution, and then towards the end of the chat a final answer is
received along with the customer expressing their sentiment (in this
case, positive).
The same interaction can happen over a voice call where a customer
service representative and a user interact to solve the issue faced by the
customer. Almost all characteristics are the same between voice calls
and chat data, except in a voice call the original data is an audio file,
which is transcribed to text first and then mined using text mining. At
the end of the customer call, the customer service representative jots
down the summary of the call. Referred to as “agent call notes,” these
notes are also mined to analyze voice calls.

SMS Data
SMS is the best way to reach 35% of the world. SMS as a channel has
one of the highest open rates (number of people who open the SMS
message to number of people who received the SMS message): 5X over
email open rates (https://blog.rebrandly.com/12-sms-
text-message-marketing-statistics). On average, a person in
the US receives 33 messages a day
(www.textrequest.com/blog/how-many-texts-people-
send-per-day/). Many app companies access customers’ SMS
messages and mine the data to improve user experiences. For instance,
apps like ReadItToMe read any SMS messages received by users while
they are driving. Truecaller reads the SMS messages and classifies them
into spam and non-spam. Walnut provides a view of users’ spending
based on the SMS messages they have received. Just by looking at only
transactional SMS data, much user information can be extracted: user’s
income, their spending, type of spending, preference for online
shopping, etc. The data source is more structured if we are only
analyzing business messages. See Figure 1-4.
Figure 1-4 A screenshot from the Walnut
(https://capitalfloat.com/walnut/) app
Businesses follow a template and are more structured. Take the
following SMS as an example. The noise in this dataset is much less.
Clear information is presented in a clear style. Although different credit
card companies can present different styles of information, it is still
easier to extract information as compared to free-flow customer text.
Mini Statement for Card ******1884.Total due Rs.
4813.70. Minimum due Rs.240.69. Payment due on 07-
SEP-19. Refer to your statement for more details.

Content Data
There is a proliferation of digital content in our lives. Online news
articles, blogs, videos, social media, and online books are key types of
content that we consume every day. On average, a consumer spends 8.8
hours consuming content digitally, per
https://cmo.adobe.com/articles/2019/2/5-consumer-
trends-that-are-shaping-digital-content-
consumption.html. The following are the key problems data
scientists need to solve to use text mining:
Content clustering (grouping similar)
Content classification
Entity recognition
Analyzing user reviews on content
Content recommendation
The other key data, like a user’s feedback on the content itself, is
more structured: number of likes, shares, clicks, time spent, and so on.
By combining the user preference data with the content data, we can
understand a lot of information about the preference of the user,
including lifestyle, life stage, hobbies, interests, and income level.

IVR Utterance Data

IVR is interactive voice response. IVR is the system that helps you
navigate various menu options in a voice call. IVR is estimated to be a
$5.5 BN market by 2023, per
www.marketsandmarkets.com/Market-
Reports/interactive-voice-response-market-
33851149.html?
gclid=Cj0KCQjwh8jrBRDQARIsAH7BsXfkISk9rnRcUPrjpBP6
qC_a3-xBFCaNzEVz4FANz5y__ZXbN7yA9ZoaAkovEALw_wcB.
Natural language-based IVR is an advanced system of IVR where a bot-
like interface helps a customer find a solution to their issue or problem.
An example of a transcript is shown in Figure 1-5.

Figure 1-5 Example IVR transcript

IVR utterances are typically short. The user knows they are talking
to a bot, so they tend to keep the sentences short. Since the utterances
are really to the point, we do not get to know other attributes or
preferences of the user other than the problem they are facing. It is a
much more structured text than the other examples you have seen so
far.

Useful Information from Data

Having described in detail the different types of data and their
applications, Figure 1-6 shows a chart that plots the useful information
value of each data source and the ease of extracting the data. The useful
information value is the value that can be extracted about a user from
the data that is useful to an organization. The ease of extracting data is
about dealing with the nature of the text corpus itself.

Figure 1-6 User information value vs. ease of extraction

© Mathangi Sri 2021
M. Sri, Practical Natural Language Processing with Python
https://doi.org/10.1007/978-1-4842-6246-7_2

2. NLP in Customer Service

Mathangi Sri1
(1) Bangalore, Karnataka, India

Customer service is a multi-billion dollar industry. The cost of bad customer experience is estimated to be a
trillion dollars globally (https://blog.hubspot.com/service/customer-service-stats).
Customer service has its early origins in the form of call centers in the 1960s. The need for customer service
grew, and today customer support has become a sizeable portion of any consumer organization. Customers
can contact organizations in a multi-modal way, through the Web, apps, voice, IVR, or calls. In this chapter,
we will look at the core problems that NLP solves in human-assisted customer service via chats and calls.
Let’s look at the data structures in the customer service industry by each of the service channels from
which text data is collected.

Voice Calls
Customers call into contact centers and a customer support agent from the company answers the customers’
queries. At the end of the call, the agent notes down the nature of the call, any tickets the agent had raised,
and the resolution. The agent may also fill out some forms that pertain to the call. Table 2-1 shows sample
data.

Table 2-1 Example of Call Data

Call id Call start Call end Agent call notes Call Agent Customer Customer survey
transferred name survey comments
rating
424798 11/12/2019 11/12/2019 customer asked No Sam
1:09:38 1:18:38 about high bills.
explained. customer
fine
450237 11/26/2019 11/26/2019 waiver fee. escalated No Kiran
8:58:58 9:13:58
794029 4/26/2019 4/26/2019 payment link sent No Karthik 3 call
8:23:52 8:34:52 disconnects..agent
helpful
249311 12/8/2019 12/8/2019 xferred to surperv Yes Megha
10:50:14 11:02:14

We analyze the agent call notes and customer survey comments. This gets then pivoted with other
structured data to derive insights about the calls.

Chats
Customers can chat with customer support on the e-commerce web page or on mobile apps. You saw an
example of a chat transcript in the last chapter. As with calls, the agents are made to fill an “agent disposition
form” at the end of an interaction. The agent disposition forms in chats are slightly longer as compared to
voice calls. The agents can multitask and hence are expected to fill in details about the chats so we get line-
level text data along with time stamps. A small example of this data is shown in Table 2-2.
Table 2-2 Example of Chat Data

Speaker Line text Time

stamp
system Thank you for choosing Best Telco. A representative will be with you shortly. 5:44:44
system You are now chatting with Max. 5:44:57
customer Himax 5:45:03
agent Thank you for contacting Best Telco. My name is Max. 5:45:14
agent Hello, I see that I’m chatting with Mrs. Sara and you’ve provided XXXXXXX as the number 5:45:26
associated with your account. Is that correct?
customer yes. that is correct. I just received email saying my service price was going up to$70/monthbut 5:45:39
on your web site, it's only$30. What's the deal with that?
This data is often analyzed with the following other data:
Metadata of chats, like average handle time, chats that got transferred to another agent, web page
where the chat happened
Metadata of agents, such as agent’s tenure and pervious performance history
Agent disposition data (as filled in by the agent), such as reason for the user’s chat, resolved or
unresolved, tickets raised, etc.
User data such as user preferences, history of user behavior data, history of user contact, etc.
User survey data such as the user rating at the end of the chat and any user feedback comments

Tickets Data
Tickets generally get created when the customer calls or chats with an agent. In few cases, one may be
created on the Web or the app directly by the customer themselves. Table 2-3 shows a quick look at ticket
data. Ticket data may be more structured than the above two channels you just examined. The unstructured
text here could be of agent notes and in some cases it will be a pretty concise form of what action was taken,
making it less useful.

Table 2-3 Example of Ticket Data

Ticket Category Subcategory Agent notes Department Status Date of ticket Date of ticket
ID opened closed
461970 Repairs Replacement sent to vendor a Merchandise Closed 10/1/2019 11:51:16 10/3/2019 11:51:16
271234 Bills Waiver closed on first CS Re- 9/28/2019 1:21:11 9/30/2019 1:21:11
call open
356827 Payment Failure called and closed Collections Closed 12/23/2019 8:06:24 12/24/2019 8:06:24
22498 Network Intermittent bad network Tech Support Open 8/10/2019 11:22:03

Email data
Email is the most cost-effective channel of customer support and is widely used. Table 2-4 shows a sample of
an Enron dataset (www.cs.cmu.edu/~./enron/) extracted from Enron CSV files.

Table 2-4 Example of an Email Dataset

Date From To Subject X-To X- X- mess

cc bcc
0 Mon, [email protected] [email protected] Tim Belden <Tim Here
14 Belden/Enron@EnronXGate> foreca
May
2001
16
1 Fri, [email protected] [email protected] Re John J Lavorato <John J Trave
4 Lavorato/ENRON@enronXgate@ENRON> have
Date From To Subject X-To X- X- mess
cc bcc
May busin
2001 meeti
13 takes
out o
trip.
Espec
you h
prepa
prese
I wou
sugge
holdi
busin
plan
meeti
here t
take a
witho
forma
busin
meeti
would
try an
some
opini
wheth
trip is
desire
neces
As far
busin
meeti
think
would
more
produ
to try
stimu
discu
acros
differ
group
what
work
what
Too o
the
prese
speak
the ot
are qu
just w
for th
turn.
meeti
might
bette
in a r
table
discu
forma
sugge
 Another Random Document on
Scribd Without Any Related Topics
tables in the patio were filling with men and women sipping their
afterdinner coffee in the cool open air.
As this was Francisco's first dinner in a hotel it might be
interesting to know what he ate. Being an Argentine, he always ate
several different kinds of meat, and began this meal with a platter of
cold meats: tongue, pressed chicken and jellied veal. Second, a
vermicelli soup with grated cheese; third, fried pejerey, the most
popular fish of the country; fourth, partridge fried in oil; fifth,
asparagus with melted butter; sixth, macaroni with tomato and
garlic sauce; seventh, roast mutton; eighth, a salad of lettuce and
tomatoes; ninth, a sweet jelly in wine sauce; tenth, fruits; and then
they adjourned to the patio for coffee.
While his uncle smoked and talked with friends, whom he had
chanced to meet, Francisco slipped away and José helped him
undress for bed, as he was very tired.
He remembered no more after José turned off the electric light
until he opened his eyes into the full glare of the sun, the next
morning. It was nine o'clock and José was laying out clean linen for
him. After a refreshing shower bath, he returned to his room to find
his rolls and coffee on a table beside his bed.
"Why, José, I'm not a lady that I must have my café in bed!"
exclaimed the lad. "Mother and the girls always do that, but I'm a
man and I want to have mine in the dining-room with Uncle Juan."
José explained that in hotels one must always take one's morning
coffee in one's rooms; and he talked on while Francisco ate and
dressed.
"El Coronel will be busy all of the day and he has placed you in
my hands. Rosario, I know like a book, and together we will see it."
"Oh! that will be great fun, José. Where shall we go first?"
"Would you like to see them load the vessels? This city is where
much of the wheat of our country is brought to be loaded into the
vessels for Europe. The river is so deep here that the largest ocean-
going vessels can come up to the docks."
They walked through crowded, busy streets until they came to a
high bluff, and from the edge of this they could look down on the
very tops of the long rows of steamships below, all being loaded
with wheat.
This was just the beginning of the busy season, for the harvest
was scarcely under way. In January and February the whole city of
Rosario would seem nothing but wheat, wheat, wheat.
Francisco saw all of this with deepest interest; he was beginning
to comprehend the resources of his own country.
They sat watching the course of the wheat bags as they shot
down the long chutes from the high bluffs to the vessels below, until
Francisco's eyes grew tired and even when he closed them he could
see long lines of bobbing bags, like yellow mice, chasing one
another into the water.
So they walked along the bluff, counting the flags of the different
nations displayed on the boats beneath them; English, French,
Italian, Dutch, German and a few that Francisco had never seen
before.
For a while they watched the lavaderas or washer-women
pounding the clothes of the city on the rocks at the edge of the
water; and spreading them on the higher rocks behind them to
bleach and dry.
Steam laundries are uncommon in South America and all of the
washing is done in this manner. The lavaderas carry the soiled linen
from the houses to the river on their heads, balancing huge bundles
as easily as though they were trifles, their arms folded across their
breasts.
As they stood watching this cleansing process Francisco spied a
raft-like boat piled high with small logs tied on securely.
 "It looks out of place here, José, among all these enormous
freight steamers. What does it carry?"
"Willow, Señorito, and see, there are others coming down the
river. It goes to Buenos Aires to be made into charcoal, the principal
fuel of that city. Great quantities of it are raised above here; it is
quick of growth and needs only to be planted so," and José
demonstrated by taking a short twig and sticking it into the earth.
"Behold! and in seven years, it is as you see it there on the rafts
ready for market. They use the twigs for making Osier baskets. But
hace calor[13] let us go to the cool shady patio of the hotel and
there I will tell you a story of some charcoal burners until the Uncle
comes."
But the Colonel reached the hotel before they did, for Francisco
must stop to see this thing and that as they sauntered along. The
mid-day heat meant little to him while so much of novelty challenged
his attention. José was always ready to answer his questions, and he
frequently drew the boy's notice to something that would escape any
one but a keen observer, and this the Indian was.
The sun was almost in midheaven, and the daily siesta was
beginning in some parts of the city. Workshops were being closed,
and under every tree some cart driver had drawn up his horse and
stretched himself on the grass under its shade; even the beggars
were curled up on the church steps fast asleep.
"Why do some of those ragged beggars wear metal badges,
José?"
"They are licensed beggars, Señorito. The city has authorized
them to beg, and when you help them you may know you are
helping no rogues."
Francisco drew his nose up into a prolonged sniff. "I believe I'm
hungry, José. What smells so good?"
"Step here on to this side street and I'll show you."
 The street was being torn up to be repaved, and the peon
workingmen at this noon interval of rest were eating their almuerzo.
Gathered in little groups, they sat around something that was
cooking and emitting odours of stewing meat, potatoes and onions.
"But how are they cooking here in the street?"
"Go closer and you can see," replied José.
Francisco walked to the curb, and looking over their backs into
the middle of one circle he saw—the stew cooking in a shovel.
"They buy these things at the market and use their street shovels
for stewpans, as you see."
"Ugh! I hope they wash them first," laughed Francisco.
They were now passing the market, an enormous affair covering
the best of a large block. But the scene was no longer animated for
the chattering and bargaining were beginning to cease; and the
merchants, themselves, were nodding over their wares.
Along the curbing were piles of merchandise; here, a stack of
peaches, pears, apricots, figs, nectarines, grapes, and plums; there,
an array of earthen ware, in curious shapes; here, a stock of
readymade clothing, aprons, trousers, ponchos[14] and shoes. The
vegetables were heaped high in piles; tomatoes, beans, lettuce,
cardon, celery, potatoes, cucumbers, and onions in long ropes, their
stems so plaited together with straw that they can be sold by the
yard; or, in that country's measure, a metro.[15]
Many of the stalls offered cooked foods; roasted partridges and
chickens; pâtes of jellied meats; cleaned and cooked armadillo,
whose meat tastes like tender roast pork. The Argentines are very
fond of them and they consume thousands every month.
Around the curbing, at one end of the market, stood great carts,
with wheels fully eight feet high. These, José told Francisco, were
the market carts that brought the produce into the city. They look
rude and cumbersome, but carry several tons and often as many as
a dozen oxen are hitched to them.
These interested Francisco but José bid him hurry as no doubt his
uncle would have breakfasted. Which, indeed, he was doing, for as
they entered the hotel Francisco caught sight of him, seated in the
long dining-room with several gentlemen; all of them, including the
Colonel, in cool looking white linen suits. Francisco joined them and
was introduced to the strangers.
They were wealthy estancieros but not Spaniards. One was an
Englishman and the other a North American, owning ranches near
Rosario, and they were negotiating with Colonel Lacevera for some
pedigreed horses which he owned.
They talked partly in Spanish and partly in English; for like most
educated Argentines, the Colonel spoke some English and
understood more. Francisco had studied English at school just as he
did French, and he was delighted to be able to understand some of
their conversation.
Before they parted, the Englishman urged Colonel Lacevera to
attend a large sale of cattle and horses which was to take place at
his estancia the next day, Sunday. Patting Francisco on the head he
added:
"Bring the Niño also, he may enjoy it."
So early the following morning José had their horses at the curb
of the hotel, saddled and ready for the three league gallop.
Francisco had not ridden often, but his enthusiasm knew no
bounds when he saw the Argentine pony that was to be his mount.
The Colonel looked at José meaningly, for he knew that this
eagerness would not outlast the long gallop.
At first they rode briskly in the cool morning air. Francisco held on
bravely, but the Colonel noticed the firm set of his lips, and that he
talked less and less as they rode on.
They were riding through beautiful country. The turf was fresh
and green in spots where the old coarse grass had been burned off
and the tender young sprouts were coming up through the rich soil.
They passed droves of several thousand sheep nibbling peacefully
on this succulent new growth. There were shepherds, with here and
there a hut made of poles covered with mud; the roof thatched with
asparta grass.
Francisco was so tired and his bones began to ache so
desperately that he ceased to show any interest in the things they
passed. Colonel Lacevera and José exchanged knowing looks, but
dared not permit Francisco to see them. When they came to one of
these rude huts his uncle said:
"Niño, would you not like to see the inside of one of these prairie
palaces?"
He admired the boy's pluck, but he feared to tax his physical
endurance more.
Francisco willingly assented, and they rode up to the door around
which a swarm of dirty, half naked children sat on the ground.
José called: "Ola!" and a copper-coloured woman appeared at the
door, dressed only in one garment, a dun-coloured chemise.
She was an Indian, and when José spoke to her in her own
tongue, asking for a drink, she pointed to the square kerosene tin
filled with water, beside which hung a gourd.
She said her husband was out with the sheep; and she had no
chairs to offer them, but they might alight and rest.
They stepped into the hut, the door of which was a horse's hide;
the floor was the hard earth; a box stood in the middle and served
as a table, while bundles of straw in the corners served as beds.
Instead of chairs there were dried skulls of oxen; their wide,
spreading horns serving as arms to these unique seats. Francisco
was glad, however, to rest his weary body within their grewsome
embrace and he sat thus for half an hour, while José watered the
horses and the Colonel talked to the children.
Francisco himself proposed that they start on, but José was
obliged to lift him into his saddle. One more league and they were in
sight of the estancia, where the sale was to be held.
The house was of the usual Spanish style of architecture, and the
many buildings grouped around it gave the place a resemblance to a
village.
Señor Stanley met them and "gave" them his house, after the
manner of all Spanish hosts, and they entered to wash and rest.
As the Señor Stanley was an Englishman, his house interested
Francisco in spite of his weariness. It was fitted with every luxury of
a high class English home; the baths being supplied with cool spring
water which flowed through them constantly. There were
handsomely furnished parlours, a well-filled library and a billiard
room. The stables were commodious and sanitary; and the tennis
courts and golf links, gardens and patios were numerous.
In the corrals they found several hundred men gathered and
there was much confusion and noise.
It was Sunday and therefore a holiday spirit pervaded everything,
for Sunday is not observed in Argentina as a day of quiet and
reverence; it is the day for sports, games and excursions. This sale
had been set for Sunday to insure a large attendance.
First, breakfast was served. Under a long arbour, formed by tall
eucalyptus trees, the table, fully a hundred feet in length, had been
set. At each place was a bunch of flowers and a bottle of native
wine.
Despite his aching body, Francisco did full justice to the soup,
barbecued meats and fowls, vegetables and fruits that were served.
But after he had eaten he crept under the shade of one of the trees
to rest.
He fell asleep and slept until his uncle wakened him at máte time.
"Hello, my boy! Slept through all of this noise? You were certainly
exhausted, for such a clatter as there has been. One hundred
thousand dollars and many pedigreed animals have changed hands,
and it wasn't done quietly either. We will have our máte and then
ride home in the cool of the evening. Come." And the Colonel helped
the stiff jointed, weary boy to his feet.

FOOTNOTES:
[13] It makes hot, literally.
[14] Blankets.
[15] A little over a yard.
 CHAPTER VII
ON THE RANCH

"What is that you have, Manuel?" cried Francisco, to one of the

peons, five days later, as he sat under an ombú tree in the garden
on his uncle's estancia, playing with some tame tierra birds, that
kept the garden clean of worms.
Manuel was one of the house peons and he had a queer looking
machine with a long snout under his arm.
"Why, this is an ant destroyer, Señorito; would you care to watch
me kill ants?"
For answer, Francisco ran eagerly to his side and the two walked
toward the peach orchard. Francisco had had five days of rest from
his tiresome ride the day of the sale, and he was now ready for any
new adventure.
They had arrived at the Tres Arroyas ranch three days before and
he had made friends with every one connected with the house and
gardens. The heat had been too great to allow of any wider
acquaintance, which would have included the gauchos, or cowboys;
at least the nearer ones, for the Tres Arroyas ranch was very large,
and Francisco never could have known them all. José had told him
that one could ride all day from the centre and not reach its
boundaries.
"Why do you use that to kill ants?" he asked of Manuel. "Our
servienta at home uses hot water when they get into the patio."
 "Ah, yes, Señorito, but these country ants come in such armies it
would take a geyser of boiling water to kill them. Now, we are here
in the orchard; you can see how they destroy things."
Curious rivulets of tawny brown ran here and there as far as the
eye could reach.
"Last spring these ants fairly cleaned our peach trees of their
tender young leaves, and it was only by continuous labour that we
exterminated them. Now, look at them! Thick as ever."
"But how can you kill millions of ants with so small a machine?"
"Well, I can't this afternoon. I brought the machine here to place
it and get it ready; then early in the morning I will tap on the iron
bars of your window and you must follow me."
It was scarcely more than dawn the next morning when Francisco
heard the gentle tapping on the rejas at his window. He had
forgotten his engagement with Manuel, and started up in
bewilderment. The sight of the peon reminded him and he hurried
into his garments and was soon with Manuel in the crisp morning air.
"A little more of the sun above the horizon and we would have
been too late for to-day," said the swarthy Spaniard, as he busied
himself lighting the machine.
"Ants are early risers, and it's only by getting up before they have
made their morning toilets that we can manage to make war on
them."
Francisco laughed at the idea of an ant bathing and dressing, and
bent over on his knees beside Manuel who was scratching a match
to light the dry rubbish in the cylindrical can, in one end of which
was a small amount of sulphur. He screwed a lid on the other end,
inserted the snout into an ant hole and with a pair of bellows he
sent the volumes of sulphurous smoke into the labyrinthine passages
of the ant houses.
 "Look, look," excitedly cried Francisco, as quantities of smoke
were seen issuing from many holes, here and there, within a radius
of several hundred yards; showing how intricate and many winding
are the underground passages of these industrious pests.
"Yes, there won't be many ants getting out to work this morning.
But in a short while they will be just as bad as ever."
They went from one part of the orchard to another until the sun
was too high, and they were obliged to stop until another morning.
Francisco learned, as they walked toward the house, that these ants
are the worst pest, excepting the locust, that the farmer has to
combat. They particularly delight in carrying away whole beds of
strawberries and they often come in armies that swarm over every
obstacle in their path.
As they entered the house, Francisco noticed that his uncle had
had café and was in his riding breeches ready for a morning gallop.
"May I go with you, Uncle Juan?" cried Francisco.
"Hey! That's spirit for you! Rode yourself to fragments a few days
ago and ready for another trial to-day. Che," clapping his hands as a
peon appeared.
"Saddle Barboza for the Señorito, inmediatamente."
Francisco gulped his café and nibbled at a biscuit, but he was too
excited to eat more.
When the horses were brought to the door, his eyes gleamed, for
he saw that the smaller horse, that was to be his to ride while he
was on the estancia, was resplendent in a new saddle, bridle and bit.
The servant brought a set of solid silver spurs and smart leather
riding boots which he assisted Francisco to put on, and which he told
him his uncle had had sent with the saddle and outfit from the city.
The stirrups were of silver, beautifully chased, and the head stall,
ornaments for the brow band which covers most of the horse's face,
and the pretel bangles that jingled across the horse's breast, were
all of the same valuable metal. It was indeed the outfit of a
gentleman, and on Barboza, the sleek bay horse, with the neat, light
hoof of the prairie steed, it seemed an equipment fit for a prince. His
uncle appeared at the mounting block and Francisco kissed him
again and again as he thanked him for the lovely gifts.
"Hey! Hey! We can't waste time thus, my boy. I am going over to
the west of the estancia to inspect some horse branding that is to
take place to-day. The mayor domo[16] will follow me later."
They cantered off across the corral and were soon on the open
plains. On and on, over the pastures, some of them red, like battle
grounds with the scarlet margarita or verbena; when again they
would reach a huge patch of white ones that looked at a distance
like snow.
"What was that, Uncle?" exclaimed Francisco, startled, as a large
bird with yellow breast and gray wings screeched across their path,
emitting a harsh cry of several syllables.
"That is the bicho-feo."[17]
"Why do they call it ugly bug? It is a bird."
"Because its cry is not unlike those words. Listen again and you
will hear how plainly he says it. It is a bird of prey and lives on
smaller birds. That bird just fluttering up out of the grass at your left
is a scissors bird."
"Oh, I know why. See how its two long tail feathers clip the air
like scissors as it flies."
They passed numbers of small gray owls; and once Francisco
spied a flock of flamingoes across the water of a small lake.
Occasionally they passed a shepherd's hut; but now they were
getting on beyond the sheep grazing pastures and great herds of
cattle came in sight.
 Francisco leaped in his saddle with joy. "Oh! Uncle, are we coming
to the cowboys?"
His boyish enthusiasm had pictured them on their native heath so
often, and now he was really to see them! He had watched them
when they came to the city on holidays and walked along the Paseo
de Julio, where the pawn shops, with their tempting offers of silver
sheathed knives, gaily striped ponchos, and silver mounted
rebenques[18] draw them as honey draws bees; but to see them on
the plains,—that was what he wanted!
He did not have to call on
his reserve of patience;
indeed, soon after his eager
question they passed a
group of them, crouched on
the ground around a fire of
dry thistles, over which
hung a can, suspended by
wire from a tripod, and
which held the water for
their morning máte. They
arose to their feet as the
Colonel galloped past and
greeted him with vivas.
"Do they often use those
murderous looking knives on
each other, Uncle?" asked
Francisco; the sight of their
weapons having subdued
his zeal somewhat. They
were rougher looking men
in their working clothes than "SOON AFTER HIS EAGER QUESTION
when they came to the city THEY PASSED A GROUP OF THEM."
dressed for a lark.
 "Seldom, Niño; unless they are intoxicated. They are not very
civilized and they have no education whatever. They fairly live on
their horses' backs and cannot be persuaded to do any work that
must be done outside their saddles."
They were, indeed, fierce in appearance. Their knee-high boots
were made of rawhide; they wore no trousers, but a striped blanket
held around the waist with a belt, then brought between the legs
and fastened again to the belt in front, formed the covering of the
lower part of the body. This is called the chirapa and when walking it
gives the wearer a bulky appearance, not unlike a Turk.
As these were peon gauchos, or low-class cowboys, they were
not so picturesque as the gentleman gaucho, who is entirely
different in appearance and character.
The mayor domo rode up to them within the first hour, and his
costume was that of the caballero class or gentleman gaucho.
He also wore the chirapa, but it was over long white cotton
trousers, the edges of which were embroidered and finished with
hand-made lace. Instead of the rawhide belt of the peon gaucho, his
was a strip of hogskin doubled, the inside forming a pocket, which
was stitched into compartments, these being made secure with
clasps made of silver coins; from all of this hung a festoon of coins
encircling the entire waist. The large clasp in the front was of solid
silver, carved to represent the crest of Argentina. Several knives
were thrust through his elaborate belt, and his riding whip was of
closely braided rawhide, with a heavy silver handle.
Francisco eyed him curiously, but with evident admiration. This
was more to his liking, and he rode between this gentleman of the
Pampas and his soldier uncle with great pride. Almost, he was
persuaded to be a gaucho, but a side glance at his idolized uncle
brought quick repentance to his heart.
How could he be so disloyal to his family traditions! A soldado,[19]
of course, that was his destiny.
 FOOTNOTES:
[16] Superintendent.
[17] Beech-o fay'-o.
[18] Riding whips.
[19] Soldier.
 CHAPTER VIII
CATTLE BRANDING

They reached the western corral about ten o'clock, and found the
branding already under way. Several dozen peon gauchos had
assembled and they had driven the horses to be branded into an
enclosure.
"See, Niño, these are all young animals; they have never had the
iron on them."
"Why do you brand them, Uncle Juan? Your estancia is so large
surely they could not stray on to a neighbour's ranch; and then the
gauchos watch them carefully?"
"Yes, but there are so many thousands that, despite the best of
care, our horses stray away occasionally. Before every yearly round-
up, we send peons to all the neighbouring ranches to gather in the
strayed ones; and if our brand is on them there is never any
question as to their owner. I am gradually having the outskirts of the
estancia enclosed in barbed wire fencing, but it is so many leagues
around that it is no easy matter. But look, see how they catch them!"
They were using the bolas, and although Francisco had often
seen them in the shop windows, he had never seen them in use.
They are an aboriginal device for lassoing cattle and horses. They
consist each of three stone balls covered with leather and all
attached to long thongs, two of which are longer than the third. The
ends of these thongs are attached together and when the gaucho
uses them he raises his hand holding these ends above his head and
whirls them around and around to gather momentum, then opening
his hand the weapon flies away to coil itself about the feet of the
animal that he wishes to lasso. These gauchos are so skilful in the
use of the bolas that their aim is unerring, and although it
sometimes bruises the captive's legs, it is a most convenient method
for catching a fleet-footed horse or cow.
When the gaucho in the
enclosure had caught a
horse by this means, he
immediately pulled it to the
ground. A peon sat on its
neck while another held it
by a rope around its fore-
legs, and a third blazed the
lines of the Tres Arroyas
brand on its hip. The mark
was in the shape of a
horseshoe, inside of which
was a cross; and at least
ten of these groups were
busy all of the time, burning
it on the young animals.
"What do you raise these
wild horses for, Uncle
Juan?" inquired Francisco,
who had not missed one
single detail of the
"BLAZED THE LINES OF THE TRES performance. "They are not
ARROYAS ON ITS HIP" fine horses like Barboza
here," and he patted his
steed's neck affectionately.
"No, they are not, by any means. These wild horses are raised for
their hides mainly, although very little of them goes to waste when
they are skinned. Look over yonder, near that cluster of mud huts,
where the hides are drying in the air and sun."
 Francisco's eyes followed the end of the silver riding whip that his
uncle used to point with, and saw tier after tier of poles, from which
were stretched horsehides to stakes in the ground below.
Turning to Don Carlos, the mayor domo, who was near-by, the
Colonel inquired the worth of the horses being branded.
"Not less than ten or twelve dollars each," answered the
superintendent. "These are very good ones. Does the Señor care to
have his breakfast now?"
For some time, Francisco had been feeling pangs of hunger. His
hurried café had not been sufficient nourishment for the long hot
ride, and now his hunger was aroused by odours that came to his
nostrils like pleasant messengers; yet, he could not see anything
cooking.
"Uncle, shall we eat out here with the gauchos?" he asked, wild-
eyed.
"Very near them anyhow, but not exactly with them. Manuel came
ahead of us to prepare our almuerzo, which is in process of cooking
over yonder behind that clump of willows. Before we eat you shall
see the gauchos eat, but I warn you it is not a prepossessing sight.
"Here, Don Carlos, have the men go to their breakfast now, the
lad wants to see their table manners."
Don Carlos rode into the corral, spoke a few words and the
branding ceased. Each man mounted his own pony, for an Argentine
cowboy never walks, be his journey ever so short. With cheers and
shouts they galloped toward the mud huts near-by.
Francisco and the Colonel followed at a more dignified pace. They
found the men gathered about in groups, squatting on the ground or
sitting on ox skulls.
The beef had been quartered and roasted on a spit over a
charcoal fire, outside one of the huts. Each man, without ceremony,
had "fallen to" and helped himself, by cutting great chunks of the
meat from the large piece on the fire.
Holding one end with his teeth and the other with his hand, each
man would sever the bite about two inches from his mouth with one
of his silver-handled belt knives.
"You see how superfluous are knives, forks and plates," said the
Colonel in an undertone to Francisco as they watched this primitive
process.
"And now for our own breakfast. I am as hollow as is the wild
pumpkin at the end of summer," and he gave a sharp blow to his
horse, another to Barboza, and they were off towards their own
waiting meal in the shadow of the willows.
Manuel had killed a small kid soon after reaching the corral, and
had roasted it on a spit in its skin over a fire of dry thistles and
charcoal. He was basting it with salt water, which he had brought in
a bottle. In the coals below were sweet potatoes roasting in their
jackets. So tempting were the combined odours of lamb and sweet
potatoes that Francisco ran to the little stream to wash himself, in
order that he might begin to appease his appetite at once.
"I never was so hungry," said he, as he took the tin plate offered
him by Manuel. "I think I could eat with my hands like the cowboys!
Do they ever eat anything but meat?"
"Seldom. They care but little for vegetables; not enough to take
the trouble of raising a few. Meat and galletas, the hard biscuit of
the Pampas, often three or four months old, is all they have besides
their máte, that they must have always.
"Que esperanza! lad, this lamb is good! It takes me back to other
days. Many times on our expeditions into the provinces have I eaten
thus."
"Tell me, do tell me of one while we eat and rest," coaxed
Francisco.
 "There were many, lad," said the Colonel, as he passed his plate
back to Manuel for another piece of the smoking, savoury lamb. "I've
never told you of the expedition of General Roca into Patagonia. I
was commanding a regiment at that time, one of the regiments that
became famous because of that remarkable undertaking.
"Patagonia is all of the southern-most part of this continent lying
between the Rio[20] Negro and the Straits of Magellan, excepting the
narrow strip between the Andes and the Pacific Ocean, which
belongs to Chile. This country is not the barren, unproductive
country now that it was before our expedition carried civilization to
its wild wastes and reclaimed those vast prairies from the Indians."
"But, Uncle Juan, what right had Argentina to take the land from
the Indians of Patagonia? They had lived there for centuries and it
was theirs."
"It is a long story, Niño, and I shall give you only the bare outline.
You see, Patagonia is a series of vast terraces from the Atlantic
Ocean to the foot of the Andes. On these well watered steppes,
Patagonian Indians, mainly the Chennas, raised their cattle, allowing
them to rove at will. But the winters there are most severe,
especially when a pampero blows; so, during the winter months,
they drove their immense droves of cattle to the northward into the
foothills of the Andes, where it was warmer. During these winter
sojourns close to the frontier of our Republic, they lived by
murdering and stealing from our settlements, and the development
of our lands was being retarded because these pioneers were
obliged to flee to the cities and leave their fields of grain and maize,
their vineyards and their cattle to the mercy of the marauders.
"Gradually the outposts of our civilization were creeping closer to
Buenos Aires, instead of extending and growing as they should. Do
you now see why we were justified in fighting them?"
"Yes, but I didn't know they had made any trouble. I supposed
they were peaceful."
 "Far from it. At last when Don Nicolas Avellaneda became
President, he sent General Roca, who was my general, and the
Minister of War, into Patagonia to exterminate these Chennas.
"It was not an easy task, for these Indians are a fierce race,
giants in size and strength. Do you know how they came by their
name, Patagonians?"
"I have never heard, it must have something to do with their feet
as 'patagon' means 'large foot.'"
"That's it exactly. Magellan, the discoverer, saw their footprints in
the sand and because of their magnitude, he believed them to be
giants, and called them that before he had ever seen them.
"Well, General Roca never knew discouragement, and he set
about their defeat by digging great trenches, twenty feet deep and
twenty feet wide, while the Indians were up in the mountains with
their herds of cattle.
"These trenches he covered with boughs, over which earth was
scattered, and when all was ready he sent us back to drive the
Chennas toward the ditches.
"It was a terrible price to pay for their cruelty, and I shudder now
as I recall that awful day; but nearly all civilization is bought with
blood, and it certainly ran in torrents then. The Indians,
unsuspecting, fell headlong, thousands of them, into the trenches,
and the few that were unhurt by the fall or by being crushed in the
trenches were made prisoners and distributed among the victorious
regiments as servants or soldiers. The women and children were
captured and sent to the cities to work.
"Ah! But those ditches! The birds, foxes, and armadillos must
have grown fat on the thousands of bodies we left on that plain."
Francisco begged for more, his eyes were ablaze and his cheeks
flushed, but the Colonel said:
 "No more of fighting, anyhow; but come here by the stream, now
that we have finished our meal, and I will tell you of some of the
animals I saw in Patagonia."
"Did you ever chase ostriches?" eagerly inquired the boy.
"Yes, yes, several times and it is great sport; and once, for three
days, I had only ostrich eggs to eat. You see, we were digging those
same trenches and could not spare many of the men for hunting. I
was ill and could not eat the army rations, so José brought me
ostrich eggs and cooked them as the Indians do—in the red-hot
coals."
"And was José with you on that expedition?" exclaimed Francisco.
"Yes, through all my campaigns he has been my body servant. It
was José who told us how the Indians catch ostriches; he had heard
it when a boy among his tribe of Araucanians."
Francisco clapped his hands in anticipation.
"A circle of fire around a great area was built and the huntsmen
remained within this circle. The ostriches and guanacos that were
thus imprisoned in the circle of fire were easy prey for they fear fire
and ran almost into our arms. Why, what's the matter, Niño?"
The interest had died out of Francisco's eyes and he sat with his
hands clasped over his knees.
"Well, Uncle Juan, I'll tell you. I'm disappointed!"
"Disappointed! How?"
"Uncle Juan, I don't think that's fair play or good sport."
"Que esperanza!" exclaimed his uncle, secretly proud of the boy's
loyalty to his conviction, but determined to draw him out on the
subject.
 "And who are you that you may sit in judgment on generals and
captains?"
"Oh! I don't think one's rank has anything to do with one's
opinions. Uncle, if a peon thinks a thing is not right he must not do
that even though the President, himself, commands him; and I don't
think hunting animals in that fashion is fair. The little English boy I
play with at school is always saying that we Spaniards are not—well,
he calls it 'sporty.' That's their English word for it. He says that the
Englishmen are the truest sports on earth and that they would never
hunt as we do."
"To a certain extent he is right, Francisco. We don't care for the
excitement of the chase merely for the excitement as they do; we
are less active in our temperament, and prefer to gain our ends with
the least expenditure of energy. I want you, above all things, my lad,
to be broad-minded, and able to see your own shortcomings, so
think this matter out and if you are convinced that we are not right
as a people, in our attitude towards sports, or anything else for that
matter, formulate your own opinions and then stick to them.
"It is through such men that all nations grow; and the men that
are able to see their national deficiencies are the great men, the
reformers, and the leaders.
"But in regard to the ostriches. How would you catch them if you
had the opportunity?"
"I should do it as the English lad tells me he saw them do it in
Chubut Territory; that's part of Patagonia, isn't it, Uncle?"
The Colonel nodded, smoking industriously.
"Well, he says the real way to catch ostriches is with the bolas.
He saw his father chase them there and he says they hunt them in
an open plain, not in a circle of fire. They give the birds an equal
chance with them for their lives, and if the ostrich can't outrun them,
then, when they are within throwing distance, they whirl the bolas
around their legs and trip them. He says it is fun to see an ostrich
run; it stretches out its long neck and with its awkward long legs
kicks up a great cloud of dust behind it. He also told me about
seeing guanacos and pumas. Did you ever hunt them, Uncle?"
"Yes, but guanacos are hard to shoot because of their keen sense
of smell, they can scent a human being over a mile away; but their
flesh is delicious, tasting much like venison.
"Have you ever seen the puma skin in the library of my city
house?"
"Yes, I have often seen it and one day I measured it; it was over
two metros in length. Are those guanaco skins in the dining-room at
the estancia—the tawny yellow ones with white spots and such deep
soft fur?"
"Yes, and the ostrich robe that your aunt uses in her carriage is
made of the breasts of young ostriches; it is as soft as down and
marked brown and white. The Patagonian Indian women often wear
them for capes, although they are very expensive.
"You know, the ostriches we have here are not the kind that
produce the long plumes worn in ladies' hats; these are called the
'rhea' and are an allied species. Speaking of skins, Francisco, I will
tell you of one that will interest you. It is a vicuña, and one of the
finest I have ever seen. It was presented to your great-grandfather,
General Lacevera, by a chief of the Incas, as a vicuña robe is worn
only by one of royal blood among the Indians. It saw service as your
great-grandfather's poncho during his remarkable career, and is now
over one hundred years old, yet it is as soft as velvet. Being one of
our family heirlooms, it shall be yours, as I have no son."
"That pleases me and I shall be very proud of it."
"As you well may be. Whatever fortunes come to you in life, Niño,
remember you are a Lacevera."
 Sleep was sweet that night, and Francisco's head was scarcely on
his pillow when guanacos, vicuñas and even gauchos were forgotten
in dreamless slumber.
 FOOTNOTE:
[20] River.
 CHAPTER IX
A SUCCESSFUL SEARCH

There was not a dull moment for Francisco during the weeks that
followed. Don Carlos, the superintendent, lived in the great house
the year through. He was a bachelor and a man of education, so
that when the Colonel came each summer he insisted that he keep
his usual quarters; for the house was very extensive and the Colonel
enjoyed his company at meals and during the long evenings.
Francisco had accompanied Don Carlos on several excursions and
once, with a tropilla of horses (eight or ten riding horses driven
loose by a peon for fresh mounts on a long journey), they had gone
on a journey of five days to a neighbouring estancia to purchase
algarroba posts for the extensive fencing that was taking place on
the Tres Arroyas ranch. This algarroba wood is like iron and under
water is almost imperishable.
They had passed by one small estancia devoted almost
exclusively to peanut culture; there were leagues and leagues of
them being raised to be shipped to the Mediterranean ports to be
made into olive oil. They had their dinner at this estancia and
Francisco ate bread made from powdered peanuts mixed with wheat
flour and he found it very delicious.
José had taken him on several fishing excursions, and once they
had hunted armadillos with small dogs. Francisco had laughed
heartily at the antics of one dog, who had almost caught the horny-
plated little animal when it suddenly rolled up into a ball, its back of
movable, bony bands enveloping it like an armour, and rolled off a
bluff over the river bank, falling fully fifty feet; while the puzzled dog
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com