An information provision system (1) comprises: a sound pickup unit (22) for picking up a guide voice and generating a voice signal (SG); a character string specifying unit (114) for specifying a registered character string that, among multiple registered character strings representing the pronunciation details of different guide voices (V), resembles a speech character string (L) representing thepronunciation details of the guide voice as analyzed by voice recognition on the voice signal (SG); and a sound-emitting unit (26) for transmitting delivery information (D), which expresses the registered character string specified by the character string specifying unit (114), to a terminal device that can present the user (U) with guidance information (G) corresponding to the delivery information D from among multiple sets of guidance information (G) corresponding to the different guide voices.