Real-Time Data Prediction Contest System for DSaaS

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing data science prediction methods, such as historical data contests, face challenges like delays, cheating, and ephemeral relevance, which hinder businesses from obtaining timely and accurate predictive models.

Innovation Solution

A real-time data prediction contest system that allows participants to build and deploy web services that respond to pre-defined questions, utilizing external data sources and enabling deferred compensation, thereby providing a platform for efficient data science as a service (DSaaS).

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If historical data prediction contests are used, then external data science expertise can be accessed, but the results are delayed and suffer from cheating issues

Engineering Contradiction:

Improveexternal data science expertiseVSAvoidtime delay in obtaining predictive models

Core Design Contradiction:

Quantity of substanceVSLoss of time

Solution Approach 1:

The system pre-registers and validates participant web services before the contest begins. Participants submit their prediction models in advance, and the system stores them ready for immediate evaluation when the contest starts, eliminating the need for real-time model submission and validation during the contest.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system extracts and separates the model validation process from the contest execution process. By validating models beforehand and storing them externally, the system removes the time-consuming validation step from the critical contest path, allowing immediate evaluation of predictions against actual outcomes.

Inventive Principle:

Principle #2Taking out (Extraction)

2Reliability

If historical data contests are used, then statistical research can be conducted, but the relevance is ephemeral and cheating occurs

Engineering Contradiction:

Improvestatistical research capabilityVSAvoidcheating and ephemeral relevance

Core Design Contradiction:

ReliabilityVSObject-affected harmful factors

Solution Approach 1:

The system transitions from static historical data contests to dynamic real-time prediction contests. The prediction targets are continuously updating data streams rather than fixed historical datasets, ensuring ongoing relevance and making cheating difficult since participants cannot access future data points that haven't occurred yet.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system implements immediate feedback mechanisms where prediction accuracy is continuously evaluated against actual outcomes as new data arrives. This real-time feedback loop maintains high standards of accuracy and discourages cheating, as poor performance is immediately apparent and affects compensation.

Inventive Principle:

Principle #23Feedback

3Adaptability or versatility

If internal staff are used for predictive analytics, then company-specific knowledge is utilized, but project backlogs cause significant delays

Engineering Contradiction:

Improvecompany-specific knowledge applicationVSAvoidspeed of delivering predictive models

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The system introduces an intermediary platform that connects company data with external data scientists. Instead of relying solely on internal staff or directly managing external consultants, the platform mediates by providing access to a broader community of data scientists who can work on company-specific problems without creating internal bottlenecks.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system creates a universal platform that can handle multiple prediction tasks simultaneously for different data streams and objectives. This multi-functional approach allows the company to address various analytical needs through a single system, increasing overall productivity while maintaining the ability to apply company-specific knowledge to each task.

Inventive Principle:

Principle #6Universality (Multi-functionality)

4Loss of time

If real-time prediction contests are implemented, then time to market is reduced, but system complexity increases

Engineering Contradiction:

Improvetime to market for predictive modelsVSAvoidsystem infrastructure complexity

Core Design Contradiction:

Loss of timeVSDevice complexity

Solution Approach 1:

The system segments the prediction contest infrastructure into modular components: participant registration, web service submission, model validation, data streaming, prediction evaluation, and compensation processing. Each module operates independently and can be developed, maintained, and scaled separately, reducing overall system complexity despite the real-time capabilities.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS11367088B2System and method for providing data science as a service

Publication Date: 2022.06.21 JPMORGAN CHASE BANK NA

AI summary

The invention relates to a computer-implemented system and method for providing data science as a service (DSaaS) using a real time data prediction contest. The method may comprise the steps of presenting a consumer interface via a network that allows a data consumer to identify a subject data source having data fields that can be predicted; and presenting a participant interface via the network that allows a participant to select a competition in which to participate and provides a web services interface enabling the participant to provide web services that can be called by the web services interface. The web services provide a prediction of the at least one data field, and the web services are permitted to utilize external data sources in generating the prediction. The prediction can then be transmitted via the network to the consumer.