This repository contains code for a project focused on mining Airbnb data from the years 2019 and 2023. The project is divided into two main parts: data exploration and query analysis, and sentiment analysis with machine learning models.
The dataset contains the following columns:
- id
- zipcode
- transit
- Bedrooms
- Beds
- Review_scores_rating
- Number_of_reviews
- Neighbourhood
- Name
- Latitude
- Longitude
- Last_review
- Instant_bookable
- Host_since
- Host_response_rate
- Host_identity_verified
- Host_has_profile_pic
- First_review
- Description
- City
- cancellation_policy
- Bed_type
- Bathrooms
- Accommodates
- Amenities
- Room_type
- Property_type
- price
- Availability_365
- Minimum_nights
In the first part of this project, we performed a series of queries on the Airbnb dataset to extract valuable insights and trends. The key tasks included:
- Query Analysis:
Executed different queries to uncover various patterns and metrics, such as the most common room types, average prices by neighborhood, and the top neighborhoods by number of reviews.
- Recommendation System:
Developed a recommendation system using Cosine Similarity to suggest similar properties to users based on specific criteria.
The second part of the project focused on analyzing the sentiment of Airbnb reviews using machine learning techniques. The steps involved:
- Annotation Process:
Annotated the comments from Airbnb reviews based on sentiment (positive, negative, neutral) using a pretrained model: bert-base-multilingual-uncased-sentiment.
- Sentiment Analysis:
Applied various machine learning algorithms to classify the sentiment of reviews:
- Support Vector Machine (SVM)
- Random Forest
- K-Nearest Neighbors (KNN)
The analysis was conducted using two different feature extraction methods:
- TF-IDF (Term Frequency-Inverse Document Frequency)
- Word Embeddings