Scoring System for Unstructured Data Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods struggle to effectively analyze and synthesize large volumes of unstructured information, such as texts and speech data, which are heterogeneous and difficult to process automatically, hindering decision-making and operational processes in various industries.

Innovation Solution

A method that assigns a score to elements in a database by converting unstructured data into structured data through a scoring process, using a dimension vector representation and interval subdivision, allowing for the extraction and representation of relevant information in structured form, enabling the synthesis of heterogeneous data types.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If unstructured data is analyzed directly, then analysis coverage is comprehensive, but analysis efficiency is low and data is not immediately usable

Engineering Contradiction:
Improveanalysis efficiencyVSAvoiddata usability
Core Design Contradiction:
ProductivityVSLoss of information

Solution Approach 1:

The patent applies preliminary action by performing pre-processing of unstructured data before analysis. The system converts unstructured data into structured format in advance, extracting relevant information and organizing it into a usable structure. This preliminary transformation enables efficient subsequent analysis while preserving all necessary information, resolving the contradiction between analysis efficiency and data usability.

Inventive Principle:
Principle #10Preliminary action

2Ease of operation

If unstructured data is converted to structured data, then data usability is improved, but processing complexity increases

Engineering Contradiction:
Improvedata usabilityVSAvoidprocessing complexity
Core Design Contradiction:
Ease of operationVSDevice complexity

Solution Approach 1:

The patent applies segmentation by breaking down the complex conversion process into distinct modules: pre-processing module, analysis module, and scoring module. Each module handles a specific aspect of the transformation from unstructured to structured data. This modular segmentation reduces processing complexity by making each step manageable and independent, while still achieving comprehensive data transformation and usability improvement.

Inventive Principle:
Principle #1Segmentation

3Speed

If automatic text analysis techniques are used, then analysis speed is improved, but accuracy of synthesizing heterogeneous information deteriorates

Engineering Contradiction:
Improveanalysis speedVSAvoidsynthesis accuracy
Core Design Contradiction:
SpeedVSMeasurement precision

Solution Approach 1:

The patent applies feedback through its multi-stage processing approach with iterative refinement. The system performs initial automatic analysis, then refines results through structured conversion and scoring mechanisms. Each stage provides feedback to improve the next stage's accuracy, maintaining high analysis speed while progressively enhancing synthesis accuracy for heterogeneous information.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent applies parameter changes by transforming data from unstructured format with variable parameters into structured format with standardized parameters. The scoring mechanism assigns numerical values to extracted information, converting qualitative heterogeneous data into quantitative structured data. This parameter standardization maintains analysis speed while significantly improving synthesis accuracy for decision-making processes.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentUS8090696B2Method and system for assigning scores to elements in a set of structured data
Publication Date: 2012.01.03 THALES SA
  • US8090696B2 patent drawing
  • US8090696B2 patent drawing
  • US8090696B2 patent drawing

AI summary

A system for implementing a scoring method, wherein the system includes at least a data analyzer configured to: determine a plurality of scoring intervals dependent upon the data to be analyzed; assign an integer score and a decimal score within the scoring intervals to each data to be analyzed, the score dependent upon a frequency of appearance; search a database for pairings of (scored element, decimal score); and generate an alert if the pairing is found in the database.