Sunday 10:15 AM–11:45 AM in Track 2

An Introduction to Sentiment Analysis of Textual Data

Fatma Tarlaci, Dhavide Aruliah

Audience level:
Novice

Description

In this tutorial, you’ll be introduced to Sentiment Analysis (SA), the extraction of subjective, affective information from text through Natural Language Processing (NLP) to enable data-driven decisions. Participants will work through a step-by-step application of SA to build a sound knowledge of its different components and an understanding of this powerful technique in various business settings

Abstract

Through this tutorial, you’ll get an introductory overview of how to make sense of a large corpus of textual data through SA. You will take advantage of open-source Python libraries (notably the Natural Language Toolkit and spaCy) to conceptualize, to design, and to implement distinct components of an SA model and workflow. After building an SA model from scratch, you’ll apply it to extract sentiments from customer review data. You’ll also get a chance to work with a pre-trained model on a corpus more representative of actual customer reviews (because they constitute an invaluable resource allowing organizations to understand quantitatively and to improve services, products, customer experiences, workforce analytics, social media monitoring processes, and, ultimately, business performance).

Approximate outline: (10 min) Setup: We’ll ensure participants have working environments and data on their machines. (20 min) Introduction to Natural Language Processing and Sentiment Analysis (15 min) Importing and preprocessing the textual data (15 min) Exercise 1: Toy example to demonstrate how sentiment analysis works (30 min) Exercise 2: Implementing a sentiment analysis model using real data

Subscribe to Receive PyData Updates

Subscribe