Retrieving design patterns by case-based reasoning and formal concept analysis

Abstract

Design patterns are applied to solve the recurring software design problems. Unfortunately, choosing appropriate design patterns for the specific design problem is considered a difficult task. A search tool is, therefore, needed to supports selection. However, the existing design pattern searching methodologies generally have problems on keywordsearch problem. In this paper, we proposed an initial model to solve the problem by using Case-Based Reasoning (CBR) and Formal Concept Analysis (FCA). For the proposed model, a case base is created to represent design patterns. FCA is used to be case organization that analyze case base for obtain knowledge embedded. The knowledge can suggest suitable indexes to use for new problem. This method is beneficial for learning index. This paper discusses the basis theoretical of representation and retrieval design patterns for this inprogress research, along with some preliminary result. © 2009 IEEE.