chapter8.tex

\documentclass{beamer}
\usepackage{latexsym} 
\usepackage{graphicx}
\usetheme{Warsaw}

\title{Chapter 5}
\subtitle{Working with Text}

\begin{document}
\maketitle

\begin{frame}
  \frametitle{}
  \begin{itemize}
  \item Natural Language Processing (NLP)
  \item Sentiment analysis (aka opinion mining)
  \item Document polarity (e.g. positive vs. negative)
  \item IMDB dataset
    \begin{itemize}
    \item 50,000 movie reviews labeled as positive/negative
    \item Positive: more than six stars on IMDB
    \item Negative: fewer than five stars on IMDB
    \end{itemize}
  \item Predict automatically whether the reviewer liked the movie
  \end{itemize}
\end{frame}

\begin{frame}
  \frametitle{Bag-of-words models}
  \begin{itemize}
  \item Idea: represent text as numerical feature vectors
  \item Create a vocabulary (alphabet) of unique tokens (e.g. words)
  \item Assign an integer index to each token
  \item Construct a sparse feature vector
  \item \href{https://github.com/rasbt/python-machine-learning-book/blob/master/code/ch08/ch08.ipynb}{CountVectorizer example}
  \end{itemize}
\end{frame}

\begin{frame}
  \frametitle{N-grams}
  \begin{itemize}
  \item Unigrams and bigrams
  \item The sun is shining
  \item Unigrams: the, sun, is, shining
  \item Bigrams: the sun, sun is, is shining
  \item CountVectorizer can extract any n-grams
  \item Tf-idf sometimes works better than row counts
  \end{itemize}
\end{frame}

\begin{frame}
  \frametitle{NLP bag-of-tricks}
  \begin{itemize}
  \item Data cleaning to remove noisy tokens (e.g. HTML tags)
  \item Stemming (e.g. running - run)
  \item Lemmatization (e.g. went - to go)
  \item Stop-word removal
  \item Open-source libraries, e.g. NLTK and OpenNLP
  \item Details in text (should be useful for projects)
  \item \href{https://github.com/rasbt/python-machine-learning-book/blob/master/code/ch08/ch08.ipynb}{90\% accurate logistic regression example}
  \item Out-of-core learning possible in scikit-learn
  \end{itemize}
\end{frame}

\begin{frame}
  \frametitle{}
  \begin{itemize}
  \item 
  \end{itemize}
\end{frame}

\end{document}