Based on JCR data, 705 samples of papers using the algorithm are screened by social survey, stratified sampling and full-text content analysis, and 16 algorithmic categories are obtained by relevant standards and expert consultation. On this basis, the usage of the algorithmic categories is counted, and the influence analysis is conducted on the number of mentions, the location of mentions and the co-occurrence based on the classified dictionary, so as to provide references for relevant study.