Analyze elder's phonetic features to find out the most efficient way to make age-robust ASR STT model