Detecting non-verbal, audible communication conveying meaning

A non-transitory computer readable medium comprising computer instructions that, when executed by at least one processor, cause the processor to:

• segment an audio into an utterance segment comprising a segment of the audio containing a party speaking;
• extract a prosodic feature and a plurality of spoken words from the utterance segment;
• apply a predictive model to the prosodic feature to determine a laughter indicator in the utterance segment;
• and
• construct an audio transcript comprising semantic characteristics and non-semantic characteristics relative to each other in the utterance segment, wherein the semantic characteristics comprise the plurality of spoken words, and wherein the non-semantic characteristics comprise the laughter indicator.

A system, comprising:

• at least one non-transitory computer readable medium;
• and
• at least one processor coupled to the at least one non-transitory computer readable medium and configured to:
• segment an audio into an utterance segment comprising a segment of the audio containing a party speaking;
• extract a prosodic feature and a plurality of spoken words from the utterance segment;
• apply a first predictive model to the prosodic feature to determine a laughter indicator in the utterance segment;
• determine an emotion for the utterance segment, wherein the computer instructions to determine the emotion for the utterance segment comprises computer instructions to:
• compare the plurality of spoken words to an emotion lexicon to identify a first emotion;
• apply a second predictive model to the prosodic feature to determine a second emotion;
• and
• apply a third predictive model to the first emotion and the second emotion to determine the emotion for the utterance segment;
• and
• construct a transcript of the audio, wherein the transcript comprises semantic characteristics and non-semantic characteristics relative to each other in the utterance segment, wherein the semantic characteristics comprise the plurality of spoken words, wherein the non-semantic characteristics comprise the laughter indicator and the emotion for the utterance segment.