Sentiment Analyzer – How Are People Feeling?

The following post is about how I developed an application named Sentiment Analyzer, utilizing Python, that aims to quantify the reaction and perception of an audience towards a subject using a popular social media platform.

The Problem – How Does My Audience Feel?

Understanding how people feel on a given subject can be challenging without a proper survey or medium to poll an audience. Gathering an appropriate audience and developing a proper survey for them are two difficult tasks all by themselves. With the rise of social media, people are expressing their thoughts and feelings publicly through platforms such as Facebook, Twitter and Instagram (to name a few).  Tapping into these data sets can provide feedback and insights that can transform organizational strategy, but also create a better understanding of our evolution as a society.

Sentiment Analyzer – My Solution

Sentiment Analyzer is the interpretation and classification of emotions (positive, negative and neutral) within text data using text analysis techniques. The text data used for this analyzer is generated by Twitter. This sentiment analyzer can help people, or businesses, identify customer sentiment toward products, brands or services in online conversations and feedback. Specifically, this application gathers tweets on a given subject, analyzes the presence of positive/negative words and returns a score for each. Using this score, positive and negative tweets are grouped into their own buckets and a percentage is reported back to the user.

About This Demonstration

The following allows a user to search a subject and return the amount of positive and negative tweets are present for this user. Please note that the gathering of tweets is limited to last 7 days, thus tweets before this time frame will not be present. If updated to pull for all historic tweets, this tool can help you understand what your audience has felt about your given subject over time.

If you want a fun example of how this application could be used, check out my audience analysis of  The Bachelor during one of it’s weekly shows.

This application will return the positive tweet and negative tweet percentage, along with an example of a positive tweet for the provided subject.

Type in the subject you are interested in and press the search icon.