Stock Pattern Assistant (SPA): A Deterministic and Explainable Framework for Structural Price Run Extraction and Event Correlation in Equity Markets

Deterministic design enables auditability: Unlike black-box predictive models or platform-dependent technical indicators, SPA uses only daily price data (OHLCV) and public events, making the entire analysis process straightforward to audit and reproduce—a critical requirement for responsible AI systems that must justify their outputs.

Explainability through factual, historical narratives: The framework generates explanations grounded in actual historical events rather than statistical correlations or opaque pattern recognition, addressing a key challenge in financial AI where transparency and factual accuracy are essential for user trust and regulatory compliance.

Clear scope limitations prevent misuse: The authors explicitly state SPA is "not a forecasting system, nor is it intended to produce trading signals," establishing appropriate guardrails that prevent the tool from being misapplied in ways that could harm users—an important responsible AI practice of defining what a system should not do.

Structural decomposition reduces noise-driven bias: By extracting monotonic price runs and filtering market noise, SPA helps separate genuine market movements from noise-driven fluctuations, reducing the risk that users might act on spurious patterns or correlations that could lead to poor financial decisions.

Modest evaluation scope warrants cautious interpretation: The evaluation covers only four equities over an unspecified period, suggesting results should be interpreted carefully and not generalized broadly—a responsible approach to communicating AI system limitations to stakeholders.