Annexe 1

This is an annexe of the publication:
Mining a Multimodal Corpus of Doctor’s Training for Virtual Patient’s Feedbacks Chris Porhet, Magalie Ochs, Jorane Saubesty, Grégoire de Montcheuil and Roxanne Bertrand International Conference on Multimodal Interaction (ICMI), 2017, Glasgow, Scotland.

Sequences

We extracted 8368 sequences from the corpus as describe in the following table:

Feedback modalitySequencesSequences length
nbr.%meanSD
Head3576434.853.37
Gaze1389174.503.40
Hand1290184.033.07
Verbal70195.393.51
Posture43154.993.27
Eyebrows32334.733.60
Smile260.43.882.25
No feedback63274.933.22

The full list of sequences in the SPMF input format are here.

Rules

We use the SMPF implementation of ERMiner to extract the rules.
Using a minimal support of 1% and a minimal confidence of 0.10, we obtains 73 rules leading to a feedback (over 715 rules).
The table list the 11 best rules:

rulesupportconfidence
doctor_hands_mvt, doctor_gaze_interlocutor, doctor_adjective ⇒ patient_head_nod850.285
doctor_head_side, doctor_gaze_interlocutor, doctor_noun ⇒ patient_head_nod870.275
doctor_verb, doctor_gaze_interlocutor, doctor_adjective ⇒ patient_head_nod850.274
doctor_adverb, doctor_gaze_interlocutor, doctor_noun ⇒ patient_head_nod1440.273
doctor_head_side, doctor_verb, doctor_gaze_interlocutor ⇒ patient_head_nod1250.272
doctor_verb, doctor_adverb, doctor_gaze_interlocutor, doctor_noun ⇒ patient_head_nod1250.271
doctor_hands_mvt, doctor_adverb, doctor_gaze_interlocutor, doctor_noun ⇒ patient_head_nod1070.270
doctor_gaze_interlocutor, doctor_adjective ⇒ patient_head_nod960.263
doctor_hands_mvt, doctor_head_side, doctor_verb, doctor_gaze_interlocutor ⇒ patient_head_nod1020.262
doctor_hands_mvt, doctor_verb, doctor_adverb, doctor_gaze_interlocutor, doctor_noun ⇒ patient_head_nod910.260
doctor_hands_mvt, doctor_verb, doctor_adverb, doctor_gaze_interlocutor ⇒ patient_head_nod1590.254

You can find the full list of rules here.