Ph.D. Dissertation Defense “A Stochastic Petri Net Reverse Engineering Methodology for Deep Understanding of Technical Documents” By Giorgia Rematska

Ph.D. Committee:  Drs. Nikolaos Bourbakis, Advisor, Bin Wang, Soon Chung, and Sukarno Mertoguno (ONR)

ABSTRACT:
Reverse Engineering has gained great attention over time and is associated with numerous different research areas. In addition, the importance of this research conducted in this thesis derives from several necessities. In particular, security analysis with learning purposes can significantly be benefited from reverse engineering. Thus, there are domains that have not yet been thoroughly investigated, like automatic reverse engineering of technical documents.
In this PhD dissertation we have developed a novel reverse engineering methodology for deep understanding of architectural description of digital hardware systems that appear in technical documents. Initially, a survey of reverse engineering of electronic or digital systems is presented. We provide a classification and summarization of research associated with this field, a maturity metric is presented to highlight weaknesses and strengths of existing methodologies and systems that are currently available.
For automatic deep understanding of technical documents, a synergistic collaboration among different modalities is proposed. Firstly, a technical document is hierarchically decomposed into two major modalities the natural language text and its images. By images we mean all the visual parts except text. Then, the natural language text is processed by a Natural Language text Processing/Understanding (NLU) methodology, and text sentences are classified into categories by utilizing a Convolutional Neural Network. Here, we consider images only the system diagrams, which are extracted and modeled by the DIM method. In particular, NLU processes the text from the document and determines the associations among the nouns and their interactions, by creating their stochastic Petri-net (SPN) graph model. DIM performs image processing and transforms the diagram in a graph form that holds all relevant information appearing in the diagram. Then, we combine (associate) these models in a synergistic way and create a SPN graph. From this SPN graph we automatically obtain the functional specifications that form the behavior of the system in a form of pseudocode. In addition, we extract a flowchart to enhance the understanding that the reader could have about the pseudocode and the hardware system as a unity.