Few-Shot Learning (FSL) aims to classify novel categories using only a handful of labeled examples per class, typically 1 to 5, which is crucial for real-world applications with scarce annotated data. Recent distribution-based approaches have demonstrated promising results by utilizing prior knowledge from base classes to estimate feature distributions for novel classes. However, most existing methods assume equal contributions from all base classes, neglecting their varying relevance. This often leads to suboptimal prior estimation and degraded performance. To tackle this issue, we propose Contribution-Aware Maximum A Posteriori (CA-MAP), a novel algorithm that explicitly quantifies the relative importance of base class prototypes when estimating class-conditional feature distributions of novel class. By integrating these weighted contributions into the distribution modeling process, CA-MAP yields more accurate and representative priors for novel categories. Extensive experiments on three widely used benchmarks, miniImageNet, CIFAR-FS, and CUB, demonstrate that CA-MAP consistently outperforms existing methods in both 1-shot and 5-shot settings. These results validate the effectiveness of the proposed contribution-aware framework for inductive FSL, offering a robust and practical solution for scenarios with limited labeled data.