The main purpose of this study is to suggest the way that develops the automatic transcription tool for foreign learner’s spoken corpus construction. The automatic transcription tool is primarily applied for efficiency. However, In terms of transcribing foreign learner’s speech is treated as the spoken corpus, transcription tools incorporate speech recognition process should be focused on the accuracy of recognition. In numerous cases, transcribing speech through the automatic tool is operated by the native speaker’s parameters. In those cases, corpus data can be established as an inappropriate form that is caused by inaccurate speech recognition based on the native speaker’s data. Therefore, the process of perceiving foreign learner’s speech should be programmed through the foreign learner’s parameters can raise the recognition rate. The foreign learner’s parameters can be composed of their own phonetic properties and frequent patterns of error.(Hongik University, Yonsei University)