This project is trying to understand consumer behavior based on the the helpfulness of the negative and positive reviews for product categories: Utilitarian and Hedonic
- In order to extract the reviews for Categories such as Music, Books etc with keywords (ex: "fiction") ordered by salesrank (or price(low to high) or inverted price (high to low)), Use the following command:
$ python test.py Books fiction salesrank
This will create a file called, Books_fiction.txt. This data will have 30 products (ordered by salesrank) and all their associated reviews.
-
A sample line of data will look like: ItemId, review_rating, #people_found_it_useful, #out_of_how_many, reviewer_id, title_of_the_review "B00CNQ7HAU","2.0","2","4","A3NNLN31LHD8Q7","So disappointing..."
-
The code in analysis.py takes the above data file as argument:
$ python analysis.py Books_fiction.txt
Output: Helpful Not Helpful Total Respondents Negative 249899 (72.29%) 95799 (27.71%) 345698 (100%) Positive 191214 (65.22%) 101974 (34.78%) 293188 (100%)