Brand Image Evaluation via Noise Removal from Social Media
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional brand evaluation methods fail to accurately assess the gap between a brand's expected image and public responses, and struggle to differentiate between noise and meaningful information across various media platforms, including social networking services.
Innovation Solution
An evaluation apparatus and method that acquires and analyzes data from multiple media sources, removes noise by identifying and substituting feeling expressions with alternative information, and calculates indices such as the 'mirror score' and 'thermo score' to evaluate brand perception and sentiment, while accounting for noise and media types.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional information analysis methods are used to collect and evaluate brand information from media, then the quantity of information can be processed, but the accuracy of evaluating gaps between expected brand image and public responses deteriorates due to noise in the acquired information
Solution Approach 1:
The patent extracts and removes noise components from acquired information by identifying expressions that do not contribute to meaningful brand evaluation. The noise removing unit specifically targets and eliminates irrelevant expressions while preserving valuable brand-related information, thereby improving evaluation accuracy without losing important data.
Solution Approach 2:
The patent changes the parameter of information quality by transforming raw, noisy information into refined, evaluation-ready information. Through the noise removal process, the system alters the state of information from contaminated to clean, enabling more precise brand image gap evaluation and public response assessment.
2Measurement precision
If feeling information is removed from provision information as noise, then evaluation accuracy improves, but loss of useful emotional context information occurs
Solution Approach 1:
The patent applies local quality by selectively removing only specific feeling information that constitutes noise in particular contexts, while preserving feeling information that provides valuable emotional context for brand evaluation. The system discerns which feeling expressions are relevant and which are noise, applying different treatment to different locations in the information.
3Adaptability or versatility
If comprehensive information from multiple media including SNS is collected, then the coverage of public responses improves, but the complexity of noise removal and evaluation increases
Solution Approach 1:
The patent implements universality by creating a unified noise removal and evaluation system that handles multiple media types (traditional media and SNS) through the same process. The evaluation apparatus applies consistent noise removal rules and evaluation criteria across all media sources, simplifying the overall complexity despite comprehensive coverage.
4Measurement precision
If post information with high matching ratio to provision information is processed, then relevant brand information is identified, but noise from repetitive expressions increases
Solution Approach 1:
The patent converts the harmful effect of repetitive expressions into a benefit by using the high matching ratio as an indicator to trigger noise removal. Repetitive expressions that initially seem harmful are actually useful signals that help identify provision information, and the system leverages this by automatically removing the repetitive noise when high matching is detected, turning the problem into a solution.
Data Source
AI summary
An evaluation apparatus includes: a provision information acquiring unit configured to acquire provision information which is associated with a target object and provided to a consumer via a medium; a post information acquiring unit configured to acquire post information posted by a poster; a noise removing unit configured to remove, as noise, at least a specific expression in the provision information from post information in which a degree of matching of the post information with the provision information is a predetermined ratio or more among pieces of post information; and an evaluation unit configured to evaluate the target object on the basis of post information whose noise has been removed by the noise removing unit.