Computing Game and Learning State in Serious Game for Learning

Abstract

In order to support the adaptive SGfL, teaching materials must be represented in game component that becomes the target of adaptivity. If adaptive architecture of the game only use game state (GS) to recognize player's state, SGfL require another indicator -learning state (LS)- to identify the learning progress. It is a necessary to formulate computational framework for both states in SGfL.The computational framework was divided into two moduls, macro-strategy and micro-strategy. Macro-strategy control the learning path based on learning map in AND-OR Graph data stucture. This paper focus on the Macro-strategy modul, that using online, direct, and centralized adaptivity method. The adaptivity in game has five components as its target. Based on those targets, eight development models of SGfL concept was enumerated. With similarity and difference analysis toward possibility of united LS and GS in computational framework to implement the nine SGfL concept into design and application, there are three groups of the development models i.e. (1) better united GS and LS, (2) must manage LS and GS as different entity, and (3) can choose whether to be united or not. In the model which is united LS with GS, computing model at the macro-strategy modul use and-or graph and forward chaining. However, in the opposite case, macro-strategy requires two intelligent computing solutions, those are and-or graph with forward chaining to manage LS collaborated with Finite State Automata to manage GS. The proposed computational framework of SGfL was resulted from the similarity and difference analysis toward all possible representations of teaching materials into the adaptive components of the game. It was not dependent of type of learning domain and also of the game genre.