Abstract: A learning support device acquires sound data; extracts a letter string representing content of an utterance for each utterance by a learner included in the sound data; identify, for each utterance, a learner who has made the utterance from among a plurality of learners based on the sound data; identify emotion information representing an emotion of the learner in the utterance, based on at least one of the sound data and data of a moving image captured together with the sound data; and outputs, for each group to which learner belongs, the plurality of pieces of emotion information in time series order with being associated with the plurality of letter strings.