The present invention relates to a text melancholyquantization method. Firstly, each emotional word taken from an emotionthesaurus is compared with a plurality of articles. Then, at least one emotional word in the plurality of articles is taken out to obtain an emotionquantization thesaurus, and a depressionvalue relative to each emotional word in emotionquantization thesaurus is set up. Finally, the depressionvalue in all emotional words in the article to be tested is computed to obtain a judgment value, providingjudgment of degree ofdepressionexpressedin thearticles. Based on the generalizedsame invention, the present invention further proposes an article melancholyquantization system based on text melancholy including: apaperinputplatform which can acceptan article input to be estimated in the degree of depression thereof atext melancholy database which can providetext melancholy and related information andan article melancholyestimationmodule which can utilize information of the text melancholy database toestimate a judgment value inputted from the paperinputplatform.
