- In this study I will work on e-commerce dataset. This dataset is very nasty. If you have a real time job regarding with data, you will see a lot of nasty datasets and you will have to work with it.
date - Timestamp of the event - not hashed
referrer_url - Referrer URL, in other words previous page - hashed
current_url - Page URL of the event - hashed
page_type - Page type of the event such as; main, productDetail, category, success etc. It is important to note that if a session contains an event with a pagetype of success, it
means that this session is completed with a purchase - not hashed
product_price - Current price of product for related page type - not hashed
cart_amount - Cart amount of user if cart is nonempty - not hashed
userid - Unique identifier of the user - hashed
sessionid - Unique identifier of the session - hashed
search_words - Searched words - hashed
old_product_price - Old price of product for related page type - not hashed
product_category - sCategory of product if it is a product page. (Array) - hashed
page_category - Category of page if it is a category page. (Array) - hashed
productid - Product id for related page type, eg. visited item, purchased items. (Array) - hashed etc...
NOTE = dataset volume is bigger than 25mb for this reason I uploaded the dataset to a web site.