Hotel Review Analysis and Prediction
Budget: $10 – $30 USD
Google Colab(please see dataset attached):
~Ensure to provide ample comments inside the code.
Data Exploration
~Load the hotel reviews data.
~Use the pandas ‘describe’ command to print out summary information about the dataset.
~Produce 2 data visualization plots.
~Describe what each of the plots indicates about the data.
Data Preprocessing
~Filter out data with rating score 0.
~Filter out data with an empty review.
Data Modeling – Predicting Hotel Ratings from Reviews
~In the next steps, you will predict ‘reviews.rating’ using ‘reviews.text’.
~Before making predictions, in a paragraph, explain what is a suitable baseline model for this problem.
~Compute the accuracy of the baseline model.
~Produce an 80-20 training-testing split of your data.
~Produce a document-term matrix for your training and testing sets.
~Use naïve bayes to fit a model to your data.
~Compute the confusion matrix on the testing set.
~In a paragraph: explain the elements of a confusion matrix, and then interpret your specific confusion matrix.
Communicating and Visualizing the Results
~Produce a plot to showcase the effectiveness of your model.
~Present a summary of your findings and their implications in a concluding paragraph.
~In a paragraph explain what steps you would take to refine your results.
~Ensure to provide ample comments inside the code.
Data Exploration
~Load the hotel reviews data.
~Use the pandas ‘describe’ command to print out summary information about the dataset.
~Produce 2 data visualization plots.
~Describe what each of the plots indicates about the data.
Data Preprocessing
~Filter out data with rating score 0.
~Filter out data with an empty review.
Data Modeling – Predicting Hotel Ratings from Reviews
~In the next steps, you will predict ‘reviews.rating’ using ‘reviews.text’.
~Before making predictions, in a paragraph, explain what is a suitable baseline model for this problem.
~Compute the accuracy of the baseline model.
~Produce an 80-20 training-testing split of your data.
~Produce a document-term matrix for your training and testing sets.
~Use naïve bayes to fit a model to your data.
~Compute the confusion matrix on the testing set.
~In a paragraph: explain the elements of a confusion matrix, and then interpret your specific confusion matrix.
Communicating and Visualizing the Results
~Produce a plot to showcase the effectiveness of your model.
~Present a summary of your findings and their implications in a concluding paragraph.
~In a paragraph explain what steps you would take to refine your results.