Finding Patterns in The Music City

Abstract

As we enter the waning side of the COVID-19 pandemic, global tourism and traveling has begun to grow again. Between 2020 and 2021, global tourism raised 4% from 400 million to 415 million. The United States saw almost 20 million international travelers during the pandemic. Including domestic travelers, tourism is a pillar for our economy. With tourism, finding a place to stay is crucial to any trip. With 7 million listings found across the world, AirBnB enables tourists to rent accommodations in specific locales. Using the Nashville 2021 dataset found at InsideAirbnb, we can apply data mining and machine learning techniques to aide the tourism industry. The dataset offers geo-location of every Airbnb in Nashville, along with the price and minimum stay length. Our first approach, we use regression to predict prices given different features about the Airbnb. The accuracy of the predictions are compared to the original, and are given a rating based on the comparison. Our second approach acts as a "related" recommender. Using clustering techniques, we explore the capability to discover similarities between Airbnbs, and suggest similar ones to the tourist. For the second data subset, sentiment analysis and word frequency is used on individual Airbnb's comments. We then build a "average sentiment" value to summarize the comments. Word frequency is used to figure out what common words are used across reviews for an Airbnb.