Book Maturity

Motivation

While there are several ways to measure the linguistic complexity of a text, I've yet to see a program that identifies a minimum age a reader needs to be for a particular book. Therefore, this project aims to meet this need.

Data Collection - Scraping Common Sense Media Book Reviews

• 0: Getting Search Pages
• 1: Getting Links to the Book Titles
• 2: Collect Every Book Review
• 3: Extract Details from Each Review

Inspect the Data

• 4: Exploratory Data Analysis

Establish the Baseline Model

• 5: Naive Baseline

XGBoost - aka "King of Kaggle"

• 6: Baseline XGBoost
• 7: XGBoost Version II
• 8: XGBoost Version III

Long Short Term Memory

• 9: Baseline LSTM
• 10: LSTM Version II
• 11: LSTM Version III
• 12: LSTM Version IV
• 13: LSTM Version V
• 14: LSTM Version VI

Computer Vision - Can you judge a book by its cover?

• 15: Get All Covers
• 16: Convolutional Neural Network (CNN)

Final Results

Model	Train MAE	Test MAE
Naive Baseline	n/a	3.27
Computer Vision	2.69	2.50
LSTM Version II	1.35	1.65
XGBoost Version III	1.50	1.62
LSTM Version VI	0.66	1.04

Presenation

Finally, to watch a presenation of this project, please click below: