Do speech registers differ in the predictability of words?
Previous research has demonstrated that language use can vary depending on the context of situation. The present paper extends this finding by comparing word predictability differences between 14 speech registers ranging from highly informal conversations to read-aloud books. We trained 14 statistical language models to compute register-specific word predictability and trained a register classifier on the perplexity score vector of the language models. The classifier distinguishes perfectly between samples from all speech registers and this result generalizes to unseen materials. We show that differences in vocabulary and sentence length cannot explain the speech register classifier’s performance. The combined results show that speech registers differ in word predictability.
- 2.Characterizing text in register analysis and natural language processing
- 4.Study 1: SLM vocabulary selection
- 4.2Results and discussion
- 5.Study 2: Training and testing of the speech register classifier
- 5.2Results and discussion
- 6.Study 3: Validation of the speech register classifier
- 6.2Results and discussion
- 7.Study 4: How much text material is needed for speech register classification?
- 7.2Results and discussion
- 8.Study 5: The sentence length confound
- 8.2Results and discussion
- 9.General discussion and conclusion