Data EngineeringNLP
From nested scraped JSON to a normalised SQLite database.
Pipeline notebook with recorded outputs; raw scraped data is not in the repository.
pandasNLTKSQLiterequests
Overview
Cleans product, colour, price, and review fields, lemmatises text with NLTK, downloads product images, and loads Product, Images, and Review tables into SQLite.
Recorded results
- Reviews loaded
- 87,398
- Exploded from nested review fields
- Source: AmazonData.ipynb, cell 28
- Image rows loaded
- 1,666
- Product images table
- Source: AmazonData.ipynb, cell 27