This paper introduces a novel approach to forecasting stock price movement patterns by combining the Rising Visibility Graph (RVG) method with the Weisfeiler-Lehman (WL) kernel. Named as RVGWL, our method transforms time series data into complex networks using RVG, capturing nuanced patterns often missed by traditional approaches. The WL kernel is then employed to quantify graph similarities and detect market trends. By integrating these techniques, we develop a robust framework for pattern recognition and prediction in financial markets. Our empirical analysis demonstrates the efficacy of our method, achieving a win rate of 92% and a mean final portfolio value of $103,192.56 from an initial investment of $100,000. This significantly outperforms the traditional Moving Average (MA) method, which recorded a 62% win rate and a mean final value of $100,925.97. These results highlight the potential of the RVGWL approach as a promising tool for financial market analysis.