Distribution Pattern Identification via Keyword Search

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Identifying the appropriate distribution pattern for test results is challenging due to the numerous possible approaches, mathematical functions, and distribution patterns available, requiring in-depth knowledge and increasing complexity.

Innovation Solution

A system and method that uses a computer to receive keywords related to the metric or test, searches documents for matching distribution patterns, applies document analysis rules, and determines confidence levels to identify expected distribution patterns, utilizing natural language processing and semantic analysis to provide accurate results.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If numerous distribution patterns and mathematical functions are available for analysis, then measurement precision and analytical capability are improved, but device complexity and difficulty of selection increase

Engineering Contradiction:
Improvedistribution pattern identification accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent introduces an intermediary system comprising a database of pre-analyzed distribution patterns and a computer program that automatically matches test results with known patterns. This intermediary layer mediates between the complex universe of distribution patterns and the user, providing automated pattern identification without requiring the user to directly navigate the complexity of numerous mathematical functions and distribution types.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If in-depth knowledge is required to identify appropriate distribution patterns, then measurement precision is improved, but ease of operation deteriorates

Engineering Contradiction:
Improvedistribution pattern selection accuracyVSAvoiduser operation simplicity
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The system enables self-service by automatically performing the complex task of distribution pattern identification. The computer program autonomously analyzes test results, searches the database for matching patterns, and identifies the appropriate distribution pattern without requiring the user to possess specialized statistical knowledge. The system serves itself by encoding expert knowledge within the automated analysis algorithm.

Inventive Principle:
Principle #25Self-service

3Adaptability or versatility

If the number of distribution patterns and metrics increases, then adaptability and analytical capability are improved, but difficulty of detecting and measuring appropriate patterns increases

Engineering Contradiction:
Improvedistribution pattern coverageVSAvoidpattern identification difficulty
Core Design Contradiction:
Adaptability or versatilityVSDifficulty of detecting and measuring

Solution Approach 1:

The patent applies preliminary action by pre-analyzing and storing numerous distribution patterns and their characteristics in a database before they are needed for actual test analysis. The system pre-processes reference data, metrics, and expected patterns so that when a test result needs analysis, the matching process can quickly retrieve and compare against pre-organized information, reducing the real-time difficulty of pattern identification despite the large number of available patterns.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS10229194B2Providing known distribution patterns associated with specific measures and metrics
Publication Date: 2019.03.12 KYNDRYL INC
  • US10229194B2 patent drawing
  • US10229194B2 patent drawing
  • US10229194B2 patent drawing

AI summary

A computer receives one or more keywords input by a user to describe a metric or a test and identifies a first document that includes at least one of the keywords. The computer identifies, in the first document, a known distribution pattern and an associated metric or measure that relate to the one or more keywords. The computer makes an electronic record of the identified known distribution pattern and associated metric or measure.