Table 2:

Results for single modality (Text, Prosodic) and multimodal approach for human generated test data.

Features	Avg. F1-score
	LSTM-l	LSTM-a
Text	0.54	0.57
Prosodic	0.42	0.42
Comb-WL	0.56	0.58
Comb-LF	0.58	0.60