An information processing apparatus is provided capable of creating images of a plurality of question sentences and answer columns from a document. A document reading unit reads a document to which marking has been applied by a plurality of different patterns and acquires document image data. A marking pattern detection unit recognizes in each different pattern the marking in a marking region to which the marking of the document image data has been applied and detects a pattern of the marking. A question sentence creation unit creates, in each different pattern, question sentence image data in which partial image data in the marking region is deleted from the document image data. An output unit outputs the question sentence image data created by the question sentence creation unit.