An image processing apparatus includes circuitry that converts user voices that are input into a string of characters reflecting a statement made with the user voices, retrieves one or more items of information related to the string, stores the information being retrieved associated with identification information indicating a retrieval time when the information is retrieved, draws a graphical image including the information being stored on a projection target image to be projected by a projector, and controls the projector to project the projection target image including the graphical image, the graphical image having a size that is determined in accordance with the identification information associated with the information included in the graphical image being projected.