Keyword Retrieval System for Group Interest Differentiation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional methods fail to effectively compare and distinguish the interest between two groups with significantly different search volumes, leading to difficulties in identifying keywords of interest specific to each group.

Innovation Solution

A computer-implemented method and system that assign keyword values based on search frequency and probability distributions between two groups, creating an output group sorted by keyword value to predict interest differences.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If the number of searches is used to compare interest between two groups, then the comparison is simple, but the accuracy of interest differentiation deteriorates when total search volumes are significantly different

Engineering Contradiction:

Improvesimplicity of comparisonVSAvoidaccuracy of interest differentiation

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The patent transforms the raw search count data into probability distributions by dividing each group's search count for a keyword by the total search count for that group. This parameter transformation normalizes the data, allowing accurate comparison of interest differentiation between groups with significantly different total search volumes. The probability distribution approach converts absolute numbers into relative proportions, resolving the contradiction between simple comparison and accurate differentiation.

Inventive Principle:

Principle #35Parameter changes

2Productivity

If raw search counts are compared directly, then the calculation is simple, but the ability to identify group-specific keywords deteriorates

Engineering Contradiction:

Improvecalculation efficiencyVSAvoidloss of interest differentiation information

Core Design Contradiction:

ProductivityVSLoss of information

Solution Approach 1:

The patent replaces the mechanical direct-subtraction approach (comparing raw search counts) with a probabilistic information processing system. By calculating probability distributions and using these to determine interest differentiation, the system preserves nuanced information about relative interest patterns that would be lost in raw count comparison. This substitution maintains computational efficiency while preventing information loss.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Device complexity

If only search count comparison is performed, then the method is simple, but the ability to predict group interest patterns deteriorates

Engineering Contradiction:

Improvemethod complexityVSAvoidpredictability of group interest

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The patent implements a feedback mechanism where the calculated probability distributions and interest differentiation metrics are used to generate predictions about group interest patterns. These predictions can then be validated against actual search behavior data, allowing the system to refine its models. This feedback loop transforms a simple static comparison into a dynamic predictive system, improving reliability while maintaining reasonable complexity.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS7958118B2Method for retrieving keyword showing difference between two groups and system for executing the method

Publication Date: 2011.06.07 NHN CORP

AI summary

A method and system for sorting out a keyword showing a difference between two groups is provided. The method includes the steps of: defining a tuple with respect to an N number of keywords and defining a data group including a group of the tuple; creating a keyword value for each of the N number of keywords by using the data group; and creating a keyword group associated with the keyword value, and creating an output group including the keyword group, wherein the two groups includes a first group and a second group.