Abstract: To quickly evaluate high-value patents, valid patents of invention and utility model owned by the tobacco industry of China were studied. An evaluation system for high-value patents was established based on the entropy weight method, which includes four dimensions: strategic value, technological value, legal value, and economic value, along with corresponding evaluation indicators for each dimension. The entropy weight method was used to calculate the weight of each indicator and finally derive a composite value for each patent. Using the Support Vector Machine (SVM) algorithm and the central difference method, a comprehensive value threshold of 0.075 1 was determined to identify high-value patents in the tobacco industry, and a total of 1 001 patents were selected. Those patents are highly concordant with the patents awarded by the industry and national certified high-value invention patents. The value of each of four dimensions across the high-value patent dataset were statistically analyzed and compared with the total patent sample set. The results indicated that the high-value patent dataset featured the highest average technological value and the lowest average total economic value, significantly outperformed the total patent sample set in all four value dimensions and the composite value. This evaluation system provides technical support for rapid and accurate screening of high-value patents.