Source Code Informal Information Analysis Using Connectionist Models

Abstract

Design recovery, a part of the reverse engineering process of a program, must supply a programmer with all the information they need to fully understand a program or a system. In this paper, a connectionist method is proposed for analyzing the informal information (comments and mnemonics) in a program that can be used for design recovery in conjunction with more traditional approaches. A connectionist model was chosen because of its properties of being robust (capable of tolerating noisy inputs), it's associative memory ability (capable of retrieving a concept given only the context of the input word that originally ftred the concept), and it's generalization ability (learn conceptually relevant micro-features of the domain). This approach uses a combination of top down domain analysis (the creation of a concept hierarchy by a domain expert, to be used in the formation of the training set) and a bottom up approach (the analysis of the informal information using the network).