2023 Brain Hack School Taiwan

---

Brainhack Taiwan 2023 Project Progess

THE Project

• Using calculated syntactical surprisals from Ngram model as a predictor in ESL speech comprehension analysis

To-do list

• Done Write the n-gram script
• 5/24 (三) 和TA開會
  1. 介紹ngram計算過程
  2. my scripts
  3. 如何融合到TRF (how to interpret the TRF results)
• lack word onset Produce TRF
• Present Project

Scripts TO-DO list

POS Trigram

• not fully Calculate the trigram count from tagged corpus(e.g. COCA) = the word frquency from the corpus
• POS tag the testing text files (=UCREL CLAWS_C7 tag)
• Preprocess the tagged testing text files
• Segment the testing txt files into trigram unit
• Calculate the probability(=surprisal) of the target POS (from the target word)

TRF Production

• Preprocess the MEG data
• Make the predictor pickle file from the surprisal of ngram (see POS Trigram part)
• lack word onset Produce the ngram TRF from the predictor
• lack word onset Analyze the ngram TRF (Statistics)