Description

The purpose of Predictable’s Product Recommendation model is to discover relationships between products through a technique called collaborative filtering. This model generates a ranked list of products to recommend to the user based on their past purchase behavior / product interest, which can be deployed for a wide variety of marketing and analytical use cases. Each product will have a score associated with each other product – the higher, the stronger the relationship between the two products.

Model Time Windows

  • PREDICTION_WINDOW: Not Applicable
  • HISTORICAL_WINDOW: 365 days

The HISTORICAL_WINDOW determines what transactions are included to calculate the relationships between the products. For a transaction to be included in the HISTORICAL_WINDOW, the purchasing customer must have made two or more purchases within the last 365 days

Data Processed:

  • Transaction Data

Results

This model delivers back a long form table that contains a row for each product pairing, with a score that represents the strength of the product pairing. Each product is has a score with every other product.

Returned Values:

  • PROD_OBS: The product that is purchased / visited
  • PROD_REC: The product to recommend, ranked in order of SCORE
  • SCORE: A number from 1-100 that indicates the strength of the relationship between PROD_OBS and PROD_REC. 100 = most similar, 1 = least similar.
  • TIME_STAMP: version of platform that trained the model
  • MODEL_VERSION: version of platform that trained the model