<![CDATA[Retail cryptocurrency investors struggle to assess portfolio risk because conventional statistical methods such as Value at Risk produce probabilistic outputs that are difficult for non-expert users to interpret. This study develops a web-based cryptocurrency wallet portfolio risk assessment system that combines Rule-Based Classification with the Herfindahl-Hirschman Index (HHI). The system evaluates each asset through four quantitative variables: market capitalization, 24-hour trading volume, 30-day price volatility, and liquidity ratio. These variables are processed by an IF-THEN rule base that produces a score from 0 to 12 and classifies assets into three risk levels, namely Low (below 5), Medium (5 to 8), and High (9 and above). At the portfolio level, HHI measures asset concentration as diversified below 0.15, moderate between 0.15 and 0.25, and concentrated at 0.25 or higher. The system covers Solana through Helius, Ethereum through Alchemy, and Bitcoin through Blockstream, with market data from CoinGecko. Testing comprised 250 automated tests on the two core modules, consisting of 87 unit tests and 163 rule-base validation tests, all passed with 100 percent code coverage on four metrics. Fourteen black-box scenarios succeeded, and Equivalence Partitioning together with Boundary Value Analysis verified the classification thresholds. Validation against an independent manual reference on 89 assets produced a 100 percent agreement level and a Cohen's Kappa of 1.0000. Evaluation of four portfolio scenarios showed that system output matched manual calculation on every component. The combination yields a transparent portfolio risk assessment presented as risk categories with traceable rule explanations.]]>
Copyrights © 2026
