By Alastair Beeson
What is Sentiment Analysis?
Sentiment Analysis is the use of Natural Language Processing, Text Analysis, Computational Linguistics, Biometrics and more to evaluate and extract insights from text.
Sentiment Analysis can be a way to turn qualitative data into quantitative data and also to process and draw insight out of a large amount of surveys or reviews without sacrificing time.
What Types of Data Can You Perform Sentiment Analysis On?
Sentiment Analysis can be performed on most kinds of text ranging from books, surveys, emails, song lyrics and more.
A popular choice for sentiment analysis are opinionated kinds of text like tweets, reviews, text messages.
Check this page to learn more about Twitter Data, get set up with their API, and create your own dataset: Twitter API and Python Scrapping Tutorial
What Are Real World Applications of Sentiment Analysis
Case: Using social media data to determine how our brand and others are perceived to design future marketing campaigns.
Case: Analyzing customer surveys and reviews to identify most desired features by our users and seeing what areas we excel versus struggle in.
Case: Using social media data to determine what types of demographics might be interested in our product.
Case: Using social media data to pre-emptively predict a protest or to predict a scandal for a celebrity.
Case: Using social media data to predict a candidate's support and chance of victory in a specific district.
What Metrics Does Sentiment Analysis Analyze?
Polarity: This means whether a piece of text's sentiment can be ascribed as "positive, negative, or neutral"
Subjectivity/Objectivity: This means whether a piece of text is more of a subjective statement i.e. emotionally charged or opinionated versus objective statement which is more impartial or factual.
Intensity: This means how intense the language is like the difference between a word like "dislike" and a word like "hate"
How Can One Perform Sentiment Analysis?
There are many dashboards for visualizing Sentiment Analysis data like Tableau, PowerBI. However, performing sentiment analysis will generally require external programs to process and analyze your data.
One of the most popular ways to perform sentiment analysis is through the Python programming language which has become the "defacto" language for data science techniques. Python has numerous data science oriented libraries including several for NLP and Sentiment Analysis including Textblob and VADER. It is pretty easy to then export your data as a CSV or JSON for use in other programs like Excel, Tableau, or even VR based Visualization applications.
Check out this page to learn more and get a tutorial: Performing Sentiment Analysis in Python