Visual teaching material that can add multi-media information by the hyperlink has so far been developed. If inappropriate information is added, however, the learner’s load is increased and the learning effect decreases. To solve this problem, we divided the information into visual-based main information that described the overview of the learning contents and text-based sub information that described the details of the contents. Moreover, to reduce missing necessary information and a load of operations, we introduced automatic reproduction of sub information by the keyword from the operation record of the visual learning material. A teaching-material-making tool was actually developed and an evaluation test was conducted in terms of learning effect by comparing with a lecture. Through this evaluation test, higher learning performance of the learning material was confirmed than that of the lecture and past visual teaching material.