Sentiment analysis

Sentiment analysis refers to a body of computational techniques to extract and categorize affect (e.g., positive or negative states) and emotions (e.g., joy, surprise) contained in text, audio, and other types of data.  It is used by businesses to analyze and understand the affect and emotions embedded in social media content such as comments, tweets, posts and blogs about their brands, products and services.

Try it!

Test sentiment analysis on a sample data set and see for yourself

Key features

Word cloud

A word cloud shows the relative frequencies of words within a corpus of textual data (e.g., user posts). Words with higher frequencies of occurrence in the corpus are placed at the center of cloud and are bigger or bolder.

Valence analysis

Valence analysis uses a dictionary to determine the polarity of words, sentences, and the entire corpus by assigning them to various affective states such as “positive,” “negative,” and “neutral.” A textual segment (e.g., a post) with more positive words than negative words is classified as a positive segment.

Emotion analysis

Emotion analysis uses a dictionary to classify textual segments into six emotion categories – anger, disgust, fear, joy, sadness and surprise.

Twitter data

Tweets can be used as data in sentiment analysis. Within Enginius, tweets can be automatically extracted for any hashtag keyword, or from any account. As a default, the program analyzes the 1,000 most relevant tweets.

Website scraping

Text from one or more web pages can also be used as a corpus for sentiment analysis. Each sentence in a corpus is considered as a post.

Case studies