关键词: 科技期刊, 表格编校, DeepSeek, 台风“杜苏芮”

Abstract: As an important expression in scientific and technical papers,the table is of great significance in improving information density,optimizing comparative analysis and enhancing academic credibility. However,there are various table editing and proofreading problems in some scientific and technical journals. Therefore,this paper summarizes the common problems according to table rules. Typhoon Doksuri researches are selected as samples to explore the identification of table editing and proofreading problems using an artificial intelligence model DeepSeek. The results show that DeepSeek has advantages in identifying the problems of quantities and units,table notes,hierarchical relationships in table headers,and data in table bodies,while it has defects in identifying the problems in formats. Due to DeepSeek's limits in understanding the contents of the whole articles and tables,it is difficult to identify the problems of self-revelation and logicality. It can be seen from this that DeepSeek plays an important role in enhancing the quality of table editing and proofreading. However,as it also has limitations,manual screening of its outputs is necessary. Strengthening the training and learning of the model serves as a valid way to increase the precision of identifying table problems and improve quality/efficiency,which can provide a reference for scientific researchers and journal editors in terms of editing and proofreading tables.

Key words: scientific and technical journals, table editing and proofreading, DeepSeek, Typhoon Doksuri