Skip to content

Latest commit

 

History

History
44 lines (37 loc) · 1.51 KB

README.md

File metadata and controls

44 lines (37 loc) · 1.51 KB
Raw

Reviews keyphrase extraction

License Python 3.7 Solvve

Description

Unsupervised keyphrase extraction from Amazon reviews for each product. For solving this problem we follow the next steps:

  • Exploratory data analysis
  • Data clean-up
  • Modeling
  • Show keyphrases for each product

Example of output

Product asin: B01AHB9CN2

  • Topic number 1: great tablet, tablet price, great tablet price, tablet love, great tablet tablet
  • Topic number 2: kindle fire, fire hd, kindle fire hd, love kindle, love kindle fire
  • Topic number 3: easy use, tablet easy, tablet easy use, easy use great, easy use tablet
  • Topic number 4: tablet great, great tablet great, great price, tablet great price, tablet great tablet
  • Topic number 5: fire hd8, amazon fire hd8, amazon fire, fire hd8 tablet, hd8 tablet

Download data

With Kaggle API download dataset:

kaggle datasets download -d datafiniti/consumer-reviews-of-amazon-products

Install all dependencies

Main dependencies:

pip install -r requirements.txt

Optional:

python -m spacy download en_core_web_lg