In embodiments, a method includes: defining a plurality of visual icons; receiving a voice message, metadata and voice tagging criteria from a first client device; converting the voice message to text; generating a summary of the converted text; generating keywords based on the summary; categorizing the voice message into categories based on the keywords and the metadata; selecting visual icons based on the categories and the tagging criteria; tagging the voice message with the selected visual icons; and sending the tagged voice message to a second client device.