Automatic speech recognition based on adaptive parameters technology in English MOOC teaching system

With the rise of MOOCs and the popularity of the internet, more and more English learners are beginning to study independently online.In order to accomplish system platform adaptation, enhance system compatibility, and increase the scoring mechanism's accuracy and dependability, this study studies a scoring approach based on adaptive parameters (AP).The system acquires the formant information of the learner's pronunciation and the standard reference pronunciation following pre-processing, FFT transformation, formant extraction, and other procedures.It contains a separate scoring parameter creation module to create adaptive parameters before speech scoring.The learner pronounces multiple voices in the scoring parameter generating module, and the expert assigns a score based on the learner's pronouncing experience.The experimental class's pre-test and post-test scores are in agreement with the design sample t-test result.This study can increase students' motivation to learn and enhance their practical English skills.