Black carbon (BC) constitutes a significant fraction of atmospheric particulate matter and plays an important role in air quality degradation and climate forcing. Despite its relevance, identifying BC emission sources remains challenging due to its structural heterogeneity and the complexity of real-world particulate matter samples. Based on the Raman spectral characteristics of black carbon, this study aims to evaluate and compare a data-driven framework for BC source identification. In the first stage, a comprehensive analysis was conducted to assess the impact of different spectral processing steps on Raman spectra of black carbon. In particular, the effects of alternative smoothing and baseline correction strategies on the stability of spectral features, the behavior of the main band parameters, and the quality of spectral fitting were examined in order to identify an optimized and reliable processing scheme.