Food Delivery Apps (DoorDash, Grubhub & UberEats) Reviews
Description
This dataset contains user-generated reviews collected from the Google Play Store for three major food delivery applications: DoorDash, Grubhub, and Uber Eats. The dataset was compiled through web scraping on November 10, 2023, and includes various attributes that capture user sentiments, textual content, and keyword-based analysis related to the Technology Acceptance Model (TAM). Data Collection Process: Source: Google Play Store user reviews Scraping Method: A Python-based web scraping script Language: English-language reviews from the US were retained Final Dataset Size (after cleaning): 1,694,581 reviews Processing Tools: Python (pandas, NRCLex, nltk, seaborn, sklearn, vaderSentiment, numpy, matplotlib) Sentiment Classification: Positive: 4-5 star ratings Neutral: 3-star ratings Negative: 1-2 star ratings Lexicon-Based Sentiment Analysis: VADER and AFINN scores were calculated to enhance sentiment classification accuracy. Dataset Columns & Descriptions: Column Name Description content The full text of the user review. score The star rating given by the user (1-5). thumbsUpCount The number of likes (thumbs up) a review received. Apps The app to which the review belongs (DoorDash, Grubhub, or Uber Eats). sentiment The sentiment category assigned based on the star rating (Positive, Neutral, Negative). afinn_score Sentiment score based on the AFINN lexicon (higher values indicate positive sentiment, lower values indicate negative sentiment). vader_neg Negative sentiment score assigned by VADER (closer to 1 indicates stronger negativity). vader_neu Neutral sentiment score assigned by VADER (closer to 1 indicates a more neutral review). vader_pos Positive sentiment score assigned by VADER (closer to 1 indicates stronger positivity). vader_compound Overall sentiment score from VADER. word_count The total number of words in the review . Ease_of_Use A value indicating whether the review contains keywords related to perceived ease of use (1 = present and positive, -1 = present and negative, 0 = absent). Intention_to_Use A value indicating whether the review contains keywords related to perceived intention to use the app (1 = present and positive, -1 = present and negative, 0 = absent). Usefulness A value indicating whether the review contains keywords related to perceived usefulness of the app (1 = present and positive, -1 = present and negative, 0 = absent). Potential Use Cases: Sentiment Analysis of user experiences with food delivery apps. User Behavior Analysis to identify factors influencing app adoption. TAM Model Validation in the context of online food delivery services. Comparative Analysis of DoorDash, Grubhub, and Uber Eats based on user feedback. Natural Language Processing (NLP) & Machine Learning Applications for review classification.