Purchase Probability Prediction Using Supervised Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Advertisers and product manufacturers face inefficiencies in identifying potential customers for targeted advertising, as existing methods often rely on blanket advertisements and lack accurate predictions of future product purchases.
Innovation Solution
A system using supervised machine learning techniques to predict the probability of entities purchasing products or product categories within a future time period, based on interest events and relevance data, generating matrices and models to estimate purchase probabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If blanket advertisements and blind contacts are used to reach potential customers, then coverage area is improved, but cost efficiency and effectiveness deteriorate
Solution Approach 1:
The patent segments the broad market into specific target groups based on purchase probability predictions. Instead of treating all potential customers equally, the system divides them into segments with different likelihoods of purchasing, allowing advertisers to focus resources on high-probability segments while maintaining broad coverage through data-driven targeting.
Solution Approach 2:
The system changes the parameter of customer selection from random or demographic-based to probability-based targeting. By calculating and using purchase probability scores as the key parameter, the system transforms blanket advertising into precision advertising, improving cost efficiency while maintaining effective coverage.
2Area of stationary object
If blanket advertisements and blind contacts are used to reach potential customers, then coverage area is improved, but effectiveness deteriorates
Solution Approach 1:
The patent segments the broad market into specific target groups based on purchase probability predictions. Instead of treating all potential customers equally, the system divides them into segments with different likelihoods of purchasing, allowing advertisers to focus resources on high-probability segments while maintaining broad coverage through data-driven targeting.
Solution Approach 2:
The system uses feedback from historical purchase data and interest events to continuously improve purchase probability predictions. By analyzing past customer behavior and feedback signals, the system refines its targeting accuracy, ensuring that blanket advertisements become increasingly effective over time while maintaining broad coverage.
3Measurement precision
If supervised machine learning techniques with matrix generation are used to predict purchase probability, then prediction accuracy is improved, but system complexity increases
Solution Approach 1:
The patent introduces matrices as an intermediary data structure between raw input data and prediction models. These matrices organize entity-product pairs and their attributes in a standardized format, simplifying the machine learning process while improving prediction accuracy. The matrix acts as a mediator that transforms complex data into a form suitable for supervised learning algorithms.
Solution Approach 2:
The system performs preliminary data processing and matrix generation before applying machine learning models. By pre-organizing data into structured matrices and preparing training datasets in advance, the system reduces the complexity of the actual prediction process while maintaining high accuracy. This preliminary action separates data preparation from model execution, making the overall system more manageable.
Data Source
AI summary
The probability that an entity will purchase a product is predicted. This prediction involves receiving input data in the form of entries each of which represents an interest event. Through a series of iterations, a final matrix is generated from the input data entries by assigning an entity-product identifier pair associated with each interest event to a different location in the matrix, along with a time identifier indicative of how far back in time the interest event occurred from a prescribed date of interest. A supervised machine learning technique is employed to create a final prediction model for each product of interest using the final matrix as input. For each product, its prediction model is applied to the input data to estimate the probability that an entity will purchase the product within a future time period.