Abstract
This paper proposes a model for a process of reconstructing text elements by focusing on a difference of input methods. With the popularization of personal computers, many people utilizes a computer for making a document. Despite such a trend, a certain number of people still makes a document by hand. The reason is that advantages of these input methods are different. To develop a novel complex input methods for mixing these advantages, we conduct a basic study to confirm characteristics of these input methods. As the first step, we examine how a input method effects a behavior of user when he or she makes a document. To focus on orders and frequencies of process components in his or her behavior during the examination, we found that typical patterns of the process can be classified into four.