Table 1.

Feature taxonomy, adapted from Voleti et al. [20]

Category	Subcategory	Feature type	Feature name, abbreviation, reference
Text-based (NLP)	Lexical features	Bag of words, vocabulary analysis	BoW, Vocab.
		Linguistic Inquiry and Word Count	LIWC [21]
		Lexical diversity	Type-Token Ratio (TTR), Moving Average TTR (MATTR), Simpson’s Diversity Index (SDI) Brunét’s Index (BI), Honoré’s Statistic (HS).
		Lexical Density	Content density (CD), Idea Density (ID), P-Density (PD).
		Part-of-Speech tagging	PoS.
	Syntactical features	Constituency-based parse tree scores	Yngve [22], Frazier [23].
		Dependency-based parse tree scores
		Speech graph	Speech Graph Attributes (SGA).
	Semantic features	Matrix decomposition methods	Latent Semantic Analysis (LSA), Principal Component Analysys (PCA).
	(Word and sentence embeddings)	Neural word/sentence embeddings	word2vec [24]
		Topic modelling	Latent Dirichlet Allocation [25].
		Psycholinguistics	Reliance on familiar words (PsyLing).
	Pragmatics	Sentiment analysis	Sent.
		Use of language UoL	Pronouns, paraphrasing, filler words (FW).
		Coherence	Coh.
Acoustic	Prosodic features	Temporal	Pause rate (PR), Phonation rate (PhR), Speech rate (SR), Articulation rate (AR). Vocalization events.
		Fundamental Frequency	F0 and trajectory.
		Loudness and energy	loud, E.
		Emotional content	emo.
	Spectral features	Formant trajectories	F1, F2, F3.
		Mel Frequency Cepstral Coefficients	MFCCs [26].
	Vocal quality	Jitter, Shimmer, harmonic-to-noise ratio	jitt, shimm, HNR.
	ASR-related	Filled pauses, repetitions, dysfluencies, hesitations. fractal dimension, entropy.	FP, rep, dys, hes, FD, entr.
		Dialogue features (i.e., Turn-Taking)	TT:avg turn length, inter-turn silences.