This is a dataset that contains 50,000 transactions and 13 features/columns, the data set is used to perform market basket analysis in association rule mining.

A random function was used during the generation and the function generates random numbers of 0's and 1's for all 13 features of each transaction.

The probability of generating a 1 is twice as high as 0, this way there will be a strong or almost-strong association between the features. 

1 means an item was purchased by the user and 0 means the item was not purchased.