方法論
この文書は、Trust Capital Index™がデジタル信頼性評価を算出するために使用する一般的なスコアリング方法論を説明しています。一般的な重み付けとディメンションの定義は、透明性を確保するために公開されています。詳細なスコアリングアルゴリズム、正規化関数、信号処理ロジックは独自のものです。
コンポジットスコア式
TCIコンポジットスコアは、それぞれ0〜100スケールに正規化された6つのサブスコアの加重平均です：
TCI = 0.25 × Visibility + 0.25 × Authority + 0.15 × Sentiment + 0.15 × Consistency + 0.10 × Risk + 0.10 × AI Perception
重み付けは固定されており、エンティティごとに調整されません。この式は、エンティティの種類、規模、業界に関係なく、すべての評価に一律に適用されます。
Realistic Score Range
Every TCI score is presented with a ±2 point realistic range to reflect the inherent variability in digital reputation data. For example, a score of 72 is displayed as 70–74, acknowledging that minor fluctuations in source availability or timing can produce slight variations in the computed result.
6つのスコアリングディメンション
可視性 (25%)
Measures the extent to which an entity is discoverable across search engines, news outlets, and digital platforms. Factors include search result count, presence on authoritative domains, and breadth of coverage across distinct sources.
権威性 (25%)
Evaluates the quality and trustworthiness of sources where the entity appears. Key signals include Wikipedia presence and article quality, domain authority of linking sites, and references from established institutional sources.
感情 (15%)
Analyzes the aggregate sentiment distribution across public mentions. Computed using keyword-based sentiment analysis and, on higher tiers, AI-powered natural language processing. Measures the ratio of positive, neutral, and negative mentions.
一貫性 (15%)
Assesses the coherence of the entity's digital identity across sources. Evaluates name consistency, description alignment, and profile uniformity. Inconsistencies may indicate fragmented or contested identity representations.
リスクエクスポージャー (10%)
Identifies adverse signals including controversies, legal proceedings, regulatory actions, and negative media patterns. Risk keywords are matched against news articles and public records to quantify exposure to reputational risk categories.
AIパーセプション (10%)
Measures how accurately AI language models represent the entity's public profile. This dimension assesses whether AI-generated descriptions align with verifiable public information, indicating the entity's digital footprint quality. Available on Strategic and Executive tiers.
正規化
各データソースからの生シグナル値は、重み付け前に共通の0〜100スケールに正規化されます。正規化は、データソース間の大きさの違いを考慮し、基礎となるデータ範囲に関係なく、各ディメンションがコンポジットスコアに比例的に貢献するようにします。
このプロセスで使用される特定の正規化関数、境界パラメータ、スケーリング係数は、独自のスコアリングモデルの一部であり、公開されていません。
TCI Rating Scale (AAA–D)
Every numeric TCI score maps mechanically to a letter-grade rating on a 7-band scale. The mapping is deterministic — no analyst override or committee review applies.
| Score | Grade | Label | Classification |
|---|---|---|---|
| 85–100 | AAA | Prime Trust | Investment Grade |
| 75–84 | AA | High Trust | Investment Grade |
| 65–74 | A | Strong Trust | Investment Grade |
| 55–64 | BBB | Moderate Trust | Investment Grade |
| 45–54 | BB | Speculative Trust | Speculative Grade |
| 35–44 | B | Weak Trust | Speculative Grade |
| 20–34 | C | Very Weak Trust | Distressed |
| 0–19 | D | Default / No Trust | Distressed |
Outlook Indicator
Computed automatically from 90-day (or 6-month) score and risk deltas.
Score Stabilization
To prevent erratic score fluctuations caused by temporary data source variations, TCI employs a score stabilization mechanism that ensures ratings evolve gradually and reflect genuine reputational changes.
Maximum Drift Cap
Scores cannot shift by more than ±8 points within any 30-day rolling window. This prevents sudden, large swings from transient data anomalies while still allowing meaningful movement when genuine reputation changes occur.
Minimum Change Threshold
Score updates smaller than ±3 points are suppressed to avoid noise. Only changes that cross the 3-point minimum threshold are reflected in the published score, ensuring that reported movements represent substantive shifts in digital reputation.
Temporal Smoothing
When monitoring is active, historical data points are weighted to produce a smooth trend line. This gives stakeholders confidence that score movements are based on sustained patterns rather than momentary data fluctuations.
TCI Transparency Standard™
The TCI Transparency Standard™ is our framework for evaluating whether an entity's public profile provides sufficient data for a reliable trust rating. A profile is classified as Certified Transparent only when all four criteria are met simultaneously.
≥ 3 Public Sources
The entity must be referenced across at least three independent public data sources (search engines, news, social media, Wikipedia, AI profiles, or verified identity).
Consistent Digital Identity
The entity must maintain a consistent identity across at least two digital platforms, with a Consistency sub-score of 50 or higher.
≥ 90-Day History
At least 90 days must have elapsed between the first and most recent analysis, ensuring the rating reflects sustained patterns rather than a single snapshot.
Stability Score > 70
The entity's score volatility must remain low, with a Stability Score exceeding 70/100, calculated from the standard deviation of historical scores.
Transparency Levels: Certified Transparent (4/4 criteria met) · High Transparency (3/4) · Partial Transparency (2/4) · Low Transparency (0–1/4). The transparency level is displayed in each report and on embeddable badges.
データソース
Trust Capital Index™は、公式APIから公開されているデータのみを使用します。制限されたプラットフォームのウェブスクレイピングは行いません。
Search Engines
TCI queries five search engines to build a comprehensive view of an entity's digital footprint, accounting for regional indexing differences and search market diversity.
DuckDuckGo
Privacy-focused search engine. Used as the primary search source for unbiased, non-personalized results across all tiers.
Google Search
World's largest search index. Country-aware queries via localized Google domains (e.g., google.de, google.co.jp) for region-specific results.
Yahoo Search
Secondary search engine providing additional coverage and cross-validation of search result counts and domain references.
Baidu Search
China's dominant search engine. Essential for entities with significant Chinese-market presence or cross-border operations.
Yandex Search
Russia and CIS region search engine. Provides coverage for entities operating in Eastern European and Central Asian markets.
News & Media
GDELT Project API
Global Database of Events, Language, and Tone. Used for news coverage analysis, sentiment distribution, and media volume tracking. Free, open access.
Google News RSS
Real-time news feed aggregation via Google News RSS endpoints. Provides recent media mentions, headline sentiment, and publication frequency analysis.
Reference & Verification
Wikipedia REST API
Official Wikimedia API with language-specific Wikipedia support. Used for entity verification, article quality assessment, and authority scoring. Queries localized Wikipedia editions based on entity country.
HTTP Link Auditing
Standard HTTP HEAD requests to verify link reachability and HTTPS adoption. No scraping — only checks HTTP response status.
Perplexity AI
AI-powered research engine used for enhanced entity profiling, social media discovery verification, and cross-referencing public information across sources.
OpenAI API (Optional)
Used on Strategic and Executive tiers for AI perception analysis, enhanced sentiment processing, executive summaries, and visual consistency analysis via GPT-4o.
Social Media Intelligence
TCI employs a 3-layer social media discovery system to identify and verify an entity's presence across 10+ social platforms including LinkedIn, X (Twitter), Facebook, Instagram, YouTube, GitHub, TikTok, Medium, and more.
Layer 1: Slug Probing
Systematic URL-pattern probing across major social platforms using common handle derivations from the entity's name. Verifies profile existence through HTTP response codes without requiring API access or authentication.
Layer 2: Cross-Platform Handle Matching
Once profiles are discovered on one platform, the system attempts to find matching handles across other platforms. This cross-referencing approach identifies consistent digital identities and improves confidence in profile attribution.
Layer 3: Perplexity AI Verification
AI-powered verification layer that uses Perplexity to confirm social media profile ownership, resolve ambiguous matches, and discover profiles that elude pattern-based detection. Provides the highest confidence level for social media attribution.
Visual Consistency Analysis
On higher tiers, TCI performs cross-platform visual identity comparison using GPT-4o vision capabilities. This analysis compares profile images, logos, and visual branding elements across discovered social media profiles and web presences.
The visual consistency engine captures profile images from each discovered platform and submits them for AI-powered comparison. The system evaluates whether the same person or brand is consistently represented, detecting mismatches that could indicate impersonation, outdated profiles, or fragmented digital identity. Results contribute to the Consistency sub-score dimension.
Country-Aware Analysis
TCI supports localized analysis for 40+ countries, adapting its data collection and processing pipeline to account for regional differences in digital presence patterns.
Localized Search Domains
Google queries are routed through country-specific domains (e.g., google.de for Germany, google.co.jp for Japan, google.com.br for Brazil) to capture region-relevant results that may not appear on google.com.
Language-Specific Wikipedia
Wikipedia queries target the appropriate language edition (e.g., de.wikipedia.org for German entities, ja.wikipedia.org for Japanese entities) to find articles that may only exist in the entity's native language.
Regional Search Engine Selection
Baidu is prioritized for Chinese entities, Yandex for Russian and CIS-region entities, and Yahoo for Japanese market coverage. This ensures that entities are evaluated using the search engines most relevant to their operating geography.
制限事項 & 免責事項
Temporal snapshot. Ratings reflect data available at the time of analysis. Digital reputations evolve, and a rating may not reflect subsequent events.
Public data only. The framework does not access private databases, social media accounts behind authentication, or proprietary datasets. Entities with limited public presence may receive lower scores due to data scarcity rather than negative reputation.
Not a credit rating. TCI ratings measure digital trust signals. They are not financial credit ratings, character assessments, or endorsements. They should not be used as the sole basis for any consequential decision.
Not legal or financial advice. The information provided by Trust Capital Index™ is for informational purposes only. It does not constitute legal advice, financial advice, or any professional recommendation. Users should consult qualified professionals for specific guidance.
Name ambiguity. Common names may produce results that conflate multiple individuals or entities. Users should review the underlying data sources to confirm relevance.
API dependencies. Rating accuracy depends on the availability and quality of data from third-party APIs. Service disruptions or data gaps may affect individual scores.
知的財産権に関する通知
Trust Capital Index™のスコアリングモデル（正規化アルゴリズム、信号処理ロジック、重み付け校正方法、コンポジットスコアリングエンジンを含む）は独自のものであり、保護された知的財産を構成します。
このページで公開されている一般的な重み付けとディメンションの定義を超えるスコアリング方法論の複製、リバースエンジニアリング、または無許可での使用は厳しく禁止されています。
© 2026 Trust Capital Index™. All rights reserved. The general framework description is provided for transparency purposes only and does not constitute a license to replicate the proprietary scoring system.
評価ティア
分析の深さはティアによって異なります。すべてのティアは同じスコアリング式と重み付けを使用します。
Preview
FreeComposite score, top risk indicators, and positive signals. Limited sub-score detail.
Starter
€19Basic report with composite score breakdown, key signals summary, and downloadable PDF. Ideal for quick personal reputation checks. 3-day access.
Core
€49Full sub-score breakdown with feature-level detail, 5 recommendations, and exportable report. Trust Badge eligibility (score ≥60). 7-day access.
Strategic
€149Includes AI Perception analysis, competitor comparison (up to 2), 30-day action plan, radar visualization, score monitoring (5 entities), and Trust Badge eligibility. 14-day access.
Executive
€299Extended competitor mapping (up to 5), AI stress-test narrative, 90-day strategic roadmap, score monitoring (20 entities), AI Trust Analyst, Trust Badge eligibility, priority processing. 30-day access.
Independent Verification Readiness
Our methodology is designed from the ground up for independent third-party verification. Every scoring decision is documented, reproducible, and auditable.
Published Formula
Complete scoring formula, dimensional weights, and composite calculation are publicly documented. No proprietary black-box components.
Full Source Attribution
Every data point in a TCI report is traceable to its original public source. Auditors can independently verify each input signal.
Reproducible Results
Given the same input data, the scoring engine produces identical results. Deterministic processing ensures audit consistency.
Audit-Ready Architecture
Immutable analysis snapshots, timestamped data collection, and comprehensive logging provide the audit trail required by Big 4 and regulatory reviewers.
Designed for Review by Leading Audit & Advisory Firms
TCI's methodology documentation, audit trails, and reproducible scoring engine are structured to meet the standards expected by global audit and advisory firms.
Methodology Whitepaper v1.0
By Octavian Badescu — Full technical whitepaper with formula, stability score, transparency standard, data architecture, and rating tiers.
The Digital Trust Capital Manifesto
By Octavian Badescu — Why trust is the new capital, and how TCI defines a new category for institutional decision-making.