Update README.md

README.md

@@ -11,6 +11,17 @@ tags:
 
 ### Dataset Description
 
-This dataset is from different Chinese poem like 《水調歌頭》and《登樓》. I chose them as my dataset is because in these poems, they have words that are being disqualified by modern age (not used anymore). For example 「邈」in s020, which means far away, but we won't use it now. Other than that, there are also words that
+This dataset is from different Chinese poem like 《水調歌頭》and《登樓》. 
+I chose them as my dataset is because in these poems, they have words that are being disqualified by modern age (not used anymore). 
+For example 「邈」in s020, which means far away, but we won't use it now. 
+Other than that, there are words that have different pronunciations, unlike how we usually talk nowadays.
+In s037, the 「華」we will pronunce it as waa4 nowadays. But in the poem it is pronunce it as faa1, 
+which is similar to the pronunciation of 「花」
 
-### Issues Encountered & Solution
+### Issues Encountered & Solution
+
+As I mentioned above, some words will have different pronunciation than its' normally, 
+I have to search on the Education Bureau corpus form each poem to make sure I have the correct pronunciations.
+After recording the audio, I used Audacity to split the long audio by sentences.
+However, Audacity will only split the audio by silences or pausing which I did when I encounter "," in the poem as well.
+Then, I double check the split and adjust them manually to make sure the audio is split according to sentences.