Sentiment analysis is a natural language processing technique that helps analyze text like data to determine positive, negative or neutral sentiments. Sentiment analysis is a powerful tool for any business looking to improve their product positioning and for startups it is particularly relevant as they base their product refinement on real customer feedback and social media discussions.
This post is an overview of basic sentiment analysis done on Google Reviews for dating apps like Tinder, Bumble and Hinge. The dataset contains information on user name, app used, app rating and review and covers data from 2016 to Jan 2022.
I. Rating analysis of the dating apps
Now, we will take a look at “Rating” given by users alongside feedback to test if majority of the customer ratings are positive or negative. As evident from the graph below, there are extreme views to the apps with a large number of customers either loving the app or hating it. Of course, this generic assessment hides the performance of individual apps in terms of ratings.
II. Classifying reviews
We have classified customer reviews into “positive” and “negative”. We classified all reviews with ‘Rating’ > 3 as +1, indicating that they are positive. All reviews with ‘Rating’ < 3 will be classified as -1. Reviews with ‘Rating’ = 3 will be dropped, because they are neutral.
As seen above, the positive sentiment word cloud has words, such as “good,” “great,” and so on.
The negative sentiment word cloud had words, such as “banned,” and “waste" and so on.
Words like 'tinder', 'matches' and 'time' were a part of both the positive and negative reviews. These can be removed to refine the word clouds for better insights.
For this and more such analyses using machine learning, get in touch with firstname.lastname@example.org.