The purpose of this study was to objectively catch the speed curve lines which occurs in the 50m sprint class, and to mining statistically about their comments so as to be accurately recognized. Participants were 35 students (19 boys and 16 girls). In order to look at the performance of this study, 50m sprint times, sprint speed, maximal speed, acceleration and comments about the speed curve were analyzed. In such analysis, Participants were separated in 50m sprint time by three groups( high, middle and lower group). The main findings are summarized as follows:１．As for the 50m sprint times, maximal speed, there were significant differences between the three groups. The high group was more higher than the middle and lower group but there were no significant differences between the three groups as for the step frequency and step length from the start to the goal.２．The words that were extracted by the text mining method for the comments for the speed curve were “speed down” and “cannot get on speed”.３．According to the words network analysis the words that were extracted inconjunction with “speed”were “think”, “speed”, “acceleration”, “down”, “10m section”,“20m section”. From these results, It was proved to say that the student could not run to speed the first half of the sprint from 10m to 20m section.