Theme Analysis
Factors: Event Descriptions and Business Categories.
In this section, we will focus on the text data. We want to understand the ongoing themes in around Yelp events and businesses. We will utilize topic modeling as the main analysis approach.
Word Clouds - What are the Top Words for Businesses and Events?
- Event descriptions
Word cloud for event descriptions. Words that stand out are New York, comedy show, join us, Edward Farrell, etc. These are the words that event planners think are the most appealing ones to show in advertising.
- Business categories
Word cloud for business categories. Words that stand out are coffee, tea, american, cocktail bars, breakfast, brunch, venues and event. We can kind of picture what New Yorkers are doing in their free time from this plot.
Topic Modeling - What Are the Common Themes Between Businesses and Events?
- Top 10 topics for the event description and business category data
Above are the topic modeling results for event descriptions and business categories. Similar topics are color coded the same way. It's not hard to see that the topics have a lot of overlaps from these two columns. In the following analysis, we will create four new columns from event descriptions: description_words, description_rows, has_url, has_image.