A method of multi-device video communication. A server receives recorded content from a multi-device video communication session and processes the recorded content to detect vocal expressions from a plurality of participants. The server generates a plurality of text elements each corresponding to one or more of the vocal expressions. The server receives at least one rating for at least one participant of the plurality of participants and generates a word cloud based on the plurality of text elements and at least in part on the at least one rating for the at least one participant.