J.A. Jacko (Ed.): Human-Computer Interaction, Part II, HCII 2009, LNCS 5611, pp. 735–744, 2009. © Springer-Verlag Berlin Heidelberg 2009 Using Pictographic Representation, Syntactic Information and Gestures in Text Entry Hamed H. Sad and Franck Poirier VALORIA lab., Université de Bretagne-Sud, Vannes, France {sad,poirier}@univ-ubs.fr Abstract. With the increasing popularity of touch screen mobile devices, it is becoming increasingly important to design fast and reliable methods for text in- put on such devices. In this work, we exploit the capabilities of those devices and a specific language model to enhance the efficiency of text entry tasks. We will distribute the roles between the user and the device in a way that allocates the tasks to the side where they can be efficiently done. The user is not a good processor of syntactic and memory retrieval operations but she/he is a highly ef- ficient processor for handling semantic and pattern recognition operations. The reverse is true for computational devices. These facts are exploited in two de- signs for the entry of common words which represent a high percentage of our written and spoken materials. A common word is typed in two or three clicks, with or without a gesture on a touch screen. Keywords: mobile text entry, pictographs, pen gestures, syntactic information. 1 Introduction Currently, the design of a usable text entry method for mobile devices is an active research area. As the device size shrinks, text entry becomes slower and error-prone [1]. In this section, we discuss the relevant characteristics of the three entities in- volved in the text entry activity: the user, the written text and the device. Clarifying these characteristics helps us to synthesis efficient and easy to use entry methods. 1.1 Text and Human Communication For humans, text is not an object in itself, but it is a mean for representing and com- municating knowledge and messages. What is mainly stored in our memory is not text but knowledge in terms of concepts, objects, scenes, functions, colors, etc. This is clear if we consider people who don’t know any writing system, they still able to store knowledge and directly communicate it. When humans communicate knowledge and information, they need some sort of coding. In the following, we discuss direct and indirect human communication and how they are related to text. By direct communication we mean person to person dialog. In our early days, we learned about our environment: objects, functions, and other semantics. For communicating these meanings, a representation was needed. The early representation we learned was an isolated utterance for each meaning (word) without any