data.md

Files inside data/local/data

• seg_p.txt and seg_q.txt contain the audio codes and their durations respectively that I used for calculating total time spoken for each speaker code.
• segments is the file that contains the Transitional speakers extracted and their audio codes and duration.
• segments_comp_p_nl and segments_comp_q_nl contain the segments of the entire JASMIN corpus for HMI and read speech respectively.
• speech_conv and speech_read contain the audio codes for each Transitional speaker for HMI and read speech respectively.
• speech_t contains the information of both speech_conv and speech_read, in a single file.
• spk2CEF_nl has information about non-natives' level of proficiency in Dutch. If there is only a speaker code mentioned, but no level, then it means the speaker is a native.
• spk2age_nl, as the name suggests, has info about the age of each speaker/
• spk2dialectregion_nl associates the speakers with the dialect they have.
• spk2gender_nl has information about speakers' gender.
• spk2group_nl has information about the age group they belong to.
  • 1: native children between 7-11 years old
  • 2: native teenagers between 12-16 years old
  • 3: non-native children between 7-16 years old
  • 4: non-native adults
  • 5: native adults above 65
• spk_t has the speaker codes for the Transitional region speakers.
• text is the file that contains the text spoken by the Transitional speakers that was used for training/testing.
• text_comp_p_nl and test_comp_q_nl contain the text spoken in the entire corpus for HMI and read respectively.
• utt2spk -> Transitional speakers file.
• utt2spk_comp_p_nl and utt2spk_comp_q_nl -> All speakers file.
• wav.scp -> Transitional audio files locations.
• wav_jas_all.scp -> same, but for the entire corpus.