Data Segment Pricing via Machine Learning

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current data marketplaces lack a technical solution for efficiently pricing arbitrary data segments, as existing pricing mechanisms are not suitable for non-standard commodities like data, which are freely replicable and lack transparency, making it difficult for data sellers to maximize revenue while ensuring buyer acceptance.

Innovation Solution

A machine learning-based system that calculates the value of data segments using features such as user traits and attributes, employing a Causal Non-linear Bandit Model to determine a revenue-maximizing offer price that aligns with market acceptance, facilitating the exchange of data segments between aggregators and consumers.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If seller-based pricing is used for data segments, then the seller can control and set pricing, but the price may be too high for consumer acceptance or too low to maximize revenue

Engineering Contradiction:

Improverevenue maximizationVSAvoidconsumer acceptance

Core Design Contradiction:

ReliabilityVSEase of operation

Solution Approach 1:

The system implements feedback loops where consumer responses to price offers are collected and used to refine pricing models. The machine learning system continuously learns from accepted and rejected offers to optimize future pricing decisions, balancing revenue maximization with consumer acceptance.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system dynamically adjusts pricing parameters based on multiple factors including data segment characteristics, consumer profiles, market conditions, and historical transaction data. This allows the pricing to adapt to different scenarios rather than using fixed seller-determined prices.

Inventive Principle:

Principle #35Parameter changes

2Adaptability or versatility

If traditional pricing mechanisms (auctions, prediction markets) are used, then pricing can be market-driven, but these mechanisms are not suitable for data commodities

Engineering Contradiction:

Improvemarket-driven pricingVSAvoidunsuitability for data

Core Design Contradiction:

Adaptability or versatilityVSObject-generated harmful factors

Solution Approach 1:

The patent replaces traditional mechanical market mechanisms (physical auctions, prediction markets) with a computational machine learning-based pricing system. This substitution adapts the pricing approach to suit the unique characteristics of data as a digital commodity that can be easily replicated and distributed.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The machine learning pricing system serves multiple functions: it estimates data segment value, generates price offers, learns from consumer responses, and adapts to different data types and consumers. This universal system replaces the need for different specialized mechanisms for different data scenarios.

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Ease of operation

If seller posted prices are used, then pricing is simple and direct, but the pricing function is not recoverable from observed data due to lack of transparency

Engineering Contradiction:

Improvepricing simplicityVSAvoidpricing transparency

Core Design Contradiction:

Ease of operationVSLoss of information

Solution Approach 1:

The machine learning model acts as an intermediary between the seller's pricing goals and the consumer's value perception. It processes observed transaction data and market signals to infer the underlying pricing function, making the pricing process transparent and recoverable from data while maintaining operational simplicity.

Inventive Principle:

Principle #24Intermediary (Mediator)

4Measurement precision

If pricing requires historical trade data, then pricing can be based on actual market behavior, but such historical data does not exist for data segments

Engineering Contradiction:

Improvepricing accuracyVSAvoiddata availability

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary actions by synthesizing initial pricing models using alternative data sources before actual trades occur. It uses simulated market data, comparable transactions, and feature-based valuation to establish baseline pricing, which is then refined as real trade data becomes available.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The pricing approach segments the problem into learnable components: data segment features, consumer characteristics, market conditions, and transaction outcomes. This segmentation allows the system to build pricing accuracy incrementally through machine learning without requiring complete historical data.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS11551194B2System to facilitate exchange of data segments between data aggregators and data consumers

Publication Date: 2023.01.10 ADOBE INC

AI summary

Techniques for exchanging data segments between data aggregators and data consumers. In an embodiment, a value of an arbitrary data segment selected by a data consumer is computed. In particular, an individual user value is calculated for each user represented in the data segment, wherein the individual user value is a weighted sum (or other function) of the one or more features of the data segment attributable to that user, plus an additive gaussian noise. The overall value of the data segment is the sum of the individual user values. An offer price for the data segment can then be calculated using the overall value. Once a request is received from the consumer to purchase the data segment at the offer price, the data segment can be exchanged between the aggregator and consumer. Thus, a data marketplace or platform for the exchange of data segments is enabled.