Sentimental Visualization: Semantic Analysis of Online Product Reviews Using Python and Tableau

The Internet has become a tool used to disseminate opinions in the form of product reviews, which significantly impact decisions to buy the product. Most previous studies have been done in this field using K-means algorithm, this study employs Python and Tableau to perform the sentiment analysis of the text as well as semantically classify the keywords into meaningful insights. Data were collected from Amazon website reviews ( n=3,413) of the product with the most reviews (Trigger Point Grid Foam Roller) in five different colors (black, camo, pink, orange, and lime) from 2018-2019. The results show that the black one has the most positive sentiment compared with the others. Moreover, SpaCy's named entity recognition revealed clusters of people, locations and purposes, or events, for the item. These results can help managers to develop the areas of a product where customers have expressed negative sentiments; moreover, this semantic classification results would help the manager to more effectively market the product.