A recommendation system is designed for users to make better choices and find what they need. At present, there are few studies on the design of recommendation systems for Chinese fonts based on emotional adjectives. Therefore, this study proposes a recommendation method for Chinese fonts based on emotional adjectives and establishes an emotional classification algorithm as well as a font assignment module, which delivers different font results according to the emotional descriptive sentences that are typed in. The method proposed by this study has demonstrated good results in preliminary experiments. A complete experiment will be conducted to verify the effect of this recommendation method for Chinese fonts.