The Fish and the Painting

Table of Contents

(First Draft)

What is this and who is it for?

About me

What other resources are there already out there?

Preface: You could, but why would you?

Research Design

Model Thinking (i.e. Think Small)

• Overview
• Defining your question
• Conceptualizing your problem
• Selecting your data (thinking about sample bias)
• Implementing your model
• Validating your model
• Discuss, discuss, discuss

Tools and Techniques

Data and code for this book

Working with metadata (hello Tables!)

Working with text data: The big picture

• What do frequency and meaning have to do with one another?

Data Preparation

• Reading in your data using the TM library
  • Normalizing your data 1: Textual normalization
  • Creating a document-term-matrix
  • Some initial metrics you can use to understand your data
  • Normalizing your data 2: Mathematical normalization 
• Reading in your data one-by-one using custom functions
• Reading in your data using BookNLP

Feature Selection and Construction

• Introducing the idea of a feature space
• Zipf’s Law and stopping to think about stopwords
• Dictionary-based approaches, including sentiment analysis
• Topic modeling
  • What is topic modeling?
  • What are some of the problems of topic modeling?
  • Chunking Texts
  • Preparing your data
  • Setting parameters of the model
  • Inspecting your model: Overview
  • Inspecting your model: Words
  • Inspecting your model: Topics
  • Assessing individual topics: Some initial diagnostics
  • Assessing Topic Stability
• Working with BookNLP Data
  • Part of speech analysis
  • Named entities (People and Places)
  • Difficulty and other readability metrics
• Social network analysis
• Advanced feature detection using machine learning

Understanding Data

Exploring

• Clustering and other unsupervised learning approaches
• Visualizing your data (learning ggplot)

Explaining

• Corpus comparison and hypothesis testing
• Comparing a single feature across multiple groups
• Multivariate regression

Predicting

• Predictive modeling and machine learning

Applications

How to think like a data-driven humanist

• Genre, or all the ways documents fall into meaningful groups
• Bias and the Social Construction of Everything
• Prestige and Social Capital
• Narrative, the most important technology ever invented
• People, or why we should call them entities
• Space, where am I?
• Readers, what are they doing?
• Careers, or the physiology of writing and creating
• Institutions, the thing we often overlook but are actually incredibly powerful shapers of our everyday reality and also supremely good at maintaining unfairness, among other things