Abstract—Nowadays is necessary to take decisions based in the knowledge obtained through advanced techniques of date analysis, decision tree is an interesting option. In this work a Rich Internet Application to visualize a decision tree in a mobile device is presented. This application lets deploy the complete tree decision and the categorization of new registers, with this tool is possible to take decisions based in the analysis of data in an extended data base. The application is developed with the framework “ZK” and requires a mobile device with internet connection capability and a web browser that support this kind of applications like: “Opera Mobile” or “Safari Mobile”. Index Terms—Decision Tree, RIA, ZK Framework, ZUML, Classification. I. INTRODUCTION The concept of Web 2.0 is meaning that a renaissance for the Internet, resulting in the emergence of innovative Web applications that have rich and interactive user interfaces. An enriched interface provides greater robustness of Web applications, since it is possible to use various types of components in the user interface, such as menus, tables, tree structures, among others, making it possible to create real applications that use the network as runtime platform, and greatly improve the user experience. A Decision Tree represents a model that consists of a set of rules to divide a large and heterogeneous population, in smaller groups and heterogeneous regarding a variable target in particular [1]. The Decision Trees are commonly used for tasks of classification and prediction. The main attraction of the methods based on trees is largely due to the fact that the Decision Trees represent rules, which can be expressed in a language that humans can understand. In this research describes the development of a Rich Internet Application using the Framework ZK. The application exploits the capabilities of advanced visualization interfaces enriched, to make it possible to visualize and work with Decision Trees on mobile devices. The application lets you view the entire tree that is being used or classify a new record, entering the values of entry and obtaining the classification given by the tree and the rule generated for this classification. II. DECISION TREES. A Decision Tree is a simple recursive structure for expressing a sequential classification process in which a record, described by a set of attributes, is assigned to one of a disjoint set of classes. [2]. It can also be seen as a model coast of a set of rules for dividing a large and heterogeneous population, in smaller groups and heterogeneous with respect to a particular target variable. Generally the target variable is a categorical variable, and the Decision Tree is used either to calculate the probability that a given record belongs to a category, or to be filed by assigning the class more probably. The Decision Trees can also be used to estimate the value of a continuous variable, although there are other techniques more suitable for this task [1]. The internal nodes in a Decision Tree involving a test for a particular attribute. Usually, proof in a node compares the value of an attribute with a constant. The leaf nodes give a classification that applies to all instances that come to this node, can also give a probability distribution over all possible classifications. A typical Decision Tree is shown in Figure 1. This represents the concept buy a computer that is, the tree tries to predict whether a customer of an electronics shop or can not buy a computer. The internal nodes are denoted by rectangles and leaf nodes ovals are denoted by [3]. Figure 1. Decision Tree for the concept to buy a computer [3]. A Rules extraction for a Tree Decision. In general, a decision tree represents a disjunction of conjunctions of restrictions on the possible values of the attributes of records. Each branch is going to the root of a tree leaf, represents a combination of such restrictions and the tree represents disjunction of those conjunctions [4]. Therefore every leaf node of the tree, it is a primitive rule production according to the form: If X 1 X 2 X 3 . . . . X n then class c Where the X i are conditions and c is the class of leaf node. There are three reasons why it should express the Decision Tree in the form of its rules of production. First, production rules are a widely-used and well-understood vehicle for representing knowledge in expert systems. Secondly, a Decision Tree can be difficult for a human expert to understand and modify, whereas the extreme modularity of production rules makes them relatively transparent. Finally, and most importantly, this transformation can improve classification performance by eliminating tests in the decision tree attributable to peculiarities of the training set, and by making it possible to combine different decision trees for the same task. [2]. Improve Decision Support using Adaptive Data Mining Alberto Ochoa-Zezzatti 1,2 , Fernando Montes 4 , Jöns Sánchez 2 , Héctor Castañeda 2 , Saúl González 1 , & Julio Ponce 3 1 Instituto de Ingeniería y Tecnología, UACJ, México. 2 CIATEC (Centro CONACYT), León de los Aldama; México. 3 Laboratorio de Inteligencia Artificial, UAA. 4 Universidad Veracruzana (Maestría en Inteligencia Artificial). E-mail: cbr_lad7@yahoo.com.mx 2009 International Conference on Electrical, Communications, and Computers 978-0-7695-3587-6/09 $25.00 © 2009 IEEE DOI 10.1109/CONIELECOMP.2009.33 61