To correctly understand why companies make charitable donations, I utilize text-based machine learning technique on news articles to identify underlying topics on what is intended by the firm when making charitable donations. Using topic modeling machine learning technique on article text, I identify four topics of donation, Altruism, Managerial Benefit, Community, and Publicity, and test hypotheses on different possible motives for donations that could be distinguished through topics identified. With the test, I find that cover-up motive and business reputation motive are main drivers of corporate charitable donation. Moreover, I also find that corporate charitable donation is associated with higher profit, reduced information asymmetry, and lower risk. Evidences from tests reveal that corporate charitable donation is mainly motivated by reputation management and is highly strategic with its potential target audiences.