Research and Perform Phonetic Error Analysis (For initial)
The fundamental elements in Hanyu Pinyin consist of initials and finals. Every Chinese syllable can be spelled with exactly one initial followed by one final, except for the special syllable “er” or when a trailing -r is considered part of a syllable [1]. Due to the difference in pronouncing cluster of letters in different languages, it is important to investigate where the non-native mispronunciation are more likely to occur in phonetic symbols of Chinese.
Phonetic error rate is defined as the total number of initial or final errors over the total number of syllables/characters. All the Hanyu Pinyin in the corpus was separated into initials and finals for each syllable. The following table shows all the initials in Hanyu Pinyin.
A Python script was written to generate phonetic error rate of initial for the entire corpus. Figure 5 shows the most mispronounced initials and their corresponding top substitution errors. We see that 6 out of 10 of these top substitution errors are related to aspiration. This trend is possibly due to the fact that while aspiration is an important feature defining many phonetic differences in Mandarin, it is not necessarily so in European languages.
Figure 5 Overall Initial Error Analysis
Reference: