JoyHIAI Jingzhi Guangnian

Evaluation Methodology

A transparent evaluation system that ensures every score stands up to scrutiny.

Current version: v2.3Last updated: 2026-07-01Review Target:19dimensions
Objective

Based on public data and scientific methods, free from commercial interests.

Transparent

The evaluation methodology and weighting system are publicly disclosed for social oversight.

Traceable

Complete data lineage and audit trail with full traceability and verification support.

Independent

Evaluation results are fully isolated from business partnerships to ensure impartiality.

Data Source

Evaluation data is drawn from five dimensions, each with defined collection frequencies and weightings to ensure comprehensive and reliable results.

Public Data Collection

40%

Corporate websites, product documentation, financial reports, patent databases, academic paper platforms, and other public channels

Daily Auto-Collection
Annual ReportPatent DatabaseProduct Release NotesAcademic Paper

Industry survey data

25%

Industry reports and market research data obtained through collaboration with third-party research institutions

Quarterly Update
IDC ReportGartner AnalysisIndustry White PaperMarket Research

Expert Assessment Data

20%

Invite AI experts to professionally evaluate and score specific dimensions.

Before each evaluation round
Technical Expert ReviewBusiness Model EvaluationMarket Outlook Assessment

Enterprise Declaration Data

10%

Qualification proofs, business data, and other submission materials proactively provided by the enterprise

Receive anytime
Business Qualification CertificatesBusiness DataCase Materials

User Feedback Data

5%

Platform user reviews, feedback on enterprises and products, and usage experience data

Continuously Collect
User RatingUser ExperienceFeature Feedback

Scoring Model

Comprehensive Score Calculation Formula

Overall Score= Σ(Dimension Score × Dimension Weights)

Quantitative Metrics

Score calculated based on Min-Max normalized raw data (0-100 points)

Qualitative indicators

Combined rating based on expert evaluation and user feedback (0-100 points)

Weight Determination Method

Step 1

AHP (Analytic Hierarchy Process)

Build the judgment matrix to determine the initial weights for each dimension.

Step 2

Delphi method calibration

Invite industry experts to conduct multi-round anonymous evaluations for weight calibration.

Step 3

Dynamic adjustment

Dynamically adjust weight allocations annually based on industry development stages.

Enterprise Evaluation Dimension Weights

Technological Innovation
25%
Commercial capabilities
20%
Market Influence
15%
Ecosystem Building Capability
15%
Organization & Talent
10%
Capital Performance
10%
Compliance & Security
5%

Product Evaluation Dimension Weights

Technical Architecture
25%
User Experience
20%
Feature Completeness
20%
AI Capabilities Depth
15%
Product-Market Fit
10%
Scalability
10%

Dimension Weights for Influencer Impact Assessment

Industry Influence
30%
Technical Vision
25%
Innovation Contribution
20%
Talent Development
15%
Social Responsibility
10%

Evaluation Period

Quarterly Review

End of each quarter

Composite Index Rankings, covering key evaluation subjects such as enterprises, products, and projects.

Published:Last week of each quarter

Annual Review

Every 12th month

Annual Leaderboard: Weighted summary of full-year data

Published:Third week of 12

Topic Evaluation

Irregular

Specialized review reports for specific tracks or trending events

Published:1-2 weeks after the event

Real-time updates

Continue

Index scores dynamically adjust based on the latest data.

Published:Critical Event Trigger

Audit Mechanism

To ensure the accuracy and credibility of review results, we have implemented a rigorous five-stage manual review process. The entire workflow typically takes 8-13 days.

Data Cleaning

1-2 daysData Engineering Team

Automated data quality checks to filter out outliers and invalid data.

Initial Review Analysis

2-3 daysAnalysis Team

Data analysts manually review collected data to ensure accuracy.

Expert Review

3-5 daysExpert Committee

3-5 industry experts independently evaluate qualitative indicators, with a weighted average score calculated.

Cross-validation

1-2 daysQuality Assurance Team

Cross-validate key data from multiple sources to ensure consistency.

Final Publish

1 daysEvaluation Committee

Publish only after the review committee approves and confirms the final results.

Business Partnership Statement

Jingzhi AI Evaluation Center hereby declares:

  • Review
    Evaluation Independence:All review results are derived from objective data and scientific methods, unaffected by any commercial partnerships.
  • Pay
    Paid promotions do not affect ratings.:Purchasing platform services (e.g., promotions, memberships) does not affect your rating score.
  • profit
    Conflict of Interest:Team members shall not participate in evaluations of companies with which they have a conflict of interest.
  • Number
    Data Transparency:The evaluation methodology and weighting system are publicly disclosed for social oversight.
  • The
    Third-party audit:Periodically invite independent third-party agencies to audit the evaluation process.

Appeal and Correction Process

If you disagree with the evaluation result, you may submit an appeal through the following process. We commit to completing the review and providing feedback within 20 business days.

1

Submit Appeal

Submit an appeal form via the platform's "Data Correction" entry. Required information includes company name, reviewed entity, objection dimension, and supporting documentation.

Submit anytime
2

Acceptance Confirmation

The Evaluation Center will confirm receipt within 5 business days, send a confirmation notice, and provide the estimated processing time.

within 5 business days
3

Verify and Process

The data team reviews the appeal and contacts the appellant for additional materials if necessary.

within 15 business days
4

Result Feedback

If correction is required, update the review data and specify the changes; otherwise, provide a detailed rationale and basis for maintaining the original result.

within 3 business days after verification

Contact Us

Data correction entry:Submit correction form online

Version History

VersionDateWhat's New
v2.32026-07-01
  • Add "AI Product Maturity Assessment" dimension
  • Increase the weight of technological innovation from 20% to 25%.
  • Optimize data cleaning algorithms
v2.22026-04-01
  • Introduce expert review mechanism
  • Add cross-validation step
  • Refine the appeal process
v2.12026-01-01
  • Add "Influence" evaluation category
  • Optimize rating normalization method
  • Increase data source transparency
v2.02025-10-01
  • Completely refactor the scoring model
  • Use AHP to determine weights.
  • Establish Expert Committee

Related Resources

Questions about the evaluation results?

We welcome all forms of supervision and feedback to help maintain the credibility of our review system.