Incorporating Fuzzy Logic into Reusable Object Management System


Saša Bošnjak, Zita Bošnjak




Reusability transfers software implementation into t he selection problem : we search across the available reusable components, described by standard attributes that capture their functional characteristic, to depict the one that is most appropriate or fulfills our software development needs to the greatest extent . One of the problems in the process of selecting from t he objects' repository lies in the fact t hat reusable components usually do not match our requirements perfectly. The central issue. therefore is, how to measure the degree of adequacy of t he chosen reusable components? In our article we propose one approach that incorporates the fuzzy sets theory and fuzzy logic into reusable objects (R-objects) management systems. The main idea is to establish one "template" R-object . based on expressed software implementation demands, and to compute its belongings to different R-objects' sets formed according to adopted R-objects classification. The measured values are in direct proportion to the correctness of selected elements from each discrimination set . Incorporating this method into t he search algorithm enables quicker access to those R-objects t hat are grouped together in a particular set as being very similar, and at the same time are most adequate for embedding into software product under development . Both the creation of "template" R-object, based on object-oriented design, and the division of the R-object repository into classification sets are described in this article.