E-mail system, which is widely used in these days, gives us the capability of reading or writing a message, making excerpts of it, and annotating to it. In this manner, various discussions or conversations are enabled using the written words.
In comparison with the text messages, voice messages are easy to input for people who don't have much skill of using a keyboard. Voice messages also contain the para-linguistic information, which enable the better discussions between the distant places. Such voice message systems are already used commonly; an automatic answering telephone is the most popular one. Voice-mail systems are also used at some offices. It is, however, very difficult to quote or annotate previous voice messages, so they are not suitable for discussions.
We designed and implemented a system called AVM (Asynchronous Virtual Meeting). It produces the effect equal to quoting and annotating other messages among speakers who participate the discussion asynchronously. The system play-backs the voice message and records the response simultaneously. It also uses the temporal information on when the voice was spoken.