Life takes us in unexpected directions if we explore the paths which open along the way. Some appeal to the two roads diverging in a forest, but for me the roads are convergent not divergent. (Poetry-Calculus Joke). My academic and professional career has led me in some strange directions which, in the end, are coming together into a harmonious whole – one which is interesting, refreshing, and comes with a whirlwind of creative endeavors.

First, and foremost I am an academic  Computer Scientist specializing in Information Assurance and Digital Forensics. I hold a B.S. and M.S. in Computer Science from the University of New Orleans. In my undergraduate work, my elective interests were primarily centered in Artificial Intelligence and Machine Learning. My masters research led me to work with our two strongest research groups in Information Assurance, and in Bioinformatics.

My work with the Bioinformatics group under Dr. Stephen Winters-Hilt, Ph.D.  involved working with hybrid Markov Model (MM) – Support Vector Machine (SVM) machine learning techniques for detecting Exon-Intron and Intron-Exon splice sites incorporating positional and relative position entropic analysis. This drew upon and extended the AI/Machine learning work I did in undergraduate studies. The work was  one facet of his nanopore detector project through Research for Children’s Diseases at Children’s Hospital. In the course of my work with the MM/SVM project, I initiated a new project to apply genetic algorithms for auto tuning SVM kernel parameters.

My work with the Information Assurance / Digital Forensics group was under the guidance of Dr. Golden Richard III, Ph.D. who served as my thesis advisor and friend during my masters studies. My studies there consisted of applying machine learning techniques I used in my research with the bioinformatics group, to the problem of Digital Forensic file reconstruction. In particular I chose to study reconstruction of textual file fragments absent header and footer information using content based analysis, natural language processing, and committee based learning. I employed the genetic algorithm auto-tuning technique I developed previously here as well, and extended the work also to go beyond the standard committee learner model, and instead used the committee learner model results to feed into another trained SVM model which showed statistically relevant improvements in the reconstruction process.

Concurrently during my masters work, I have worked professionally over the last 5 years as the IT director of a litigation support company. My department had multiple areas of responsibility, and was tasked, under my leadership and direction, with oversight and administration of a network of sister companies. In the course of our work we centrally handled a diverse network topology spanning 3 states and 11 offices. This has given me an extremely in-depth view of internal corporate Information Technology including the server architectures, network topologies both internal LAN, and site-to-site WAN, end-user security concerns, document retention policies, backup plans, and disaster recovery.

Beyond the more mundane IT role my department oversaw, I personally directed the development of our Electronic Discovery offerings in response to the changes to the Federal Rules of Civil Procedure in 2006 regarding eDiscovery. This development continues as we expand service offerings to meet the changing legal landscape.  I have been accepted as an expert in the forensic investigation of IT systems by the Louisiana courts. I have been involved in recent novel discovery disputes in which motions to compel resulted in physical access to internal corporate systems to oversee and advise on data acquisition protocols, and search query sufficiency. I also conduct classic forensic acquisition, drive imaging, and investigations.

I am currently a PhD student in Engineering and Applied Sciences at the University of New Orleans. I am presently working on my dissertation which includes case studies from direct experience with large scale corporate eDiscovery, field research on site at several large corporations observing their document retention protocols and the problems they present for eDiscovery, and the inherent problems associated with eDiscovery protocols and searches being crafted and executed by individuals who are not sufficiently knowledgeable to do so. There are several threads of court decisions which support my observations and hypothesis which I will explore in depth. The work will ultimately examine the human bottleneck I’ve spoken of previously in a presentation to the American Academy of Forensic Sciences (AAFS) in 2010, what industry is currently doing, and examining the effectiveness of current information retrieval techniques for automatic-coding of a document corpus to discern privilege, and responsive document types. The ultimate goal is to take established practice, determine its efficacy and craft open source tools and protocols to impose scientific rigor on eDiscovery practices.

As of Fall 2010, I am also a law student at Tulane Law School in New Orleans. Though my PhD will be completed before then, I will be completing my J.D. in 2013. The law and digital forensics are innately intertwined, and I hope to explore how the two disciplines will complement each other. My current primary interest areas are Civil Procedure as it relates to the discovery process important to law in general, Intellectual Property, and technology law – and of course mediation/ADR for eDiscovery disputes. This is a new venture, so I do not prejudge where the path may lead, but if you are interested in collaboration or, ahem, looking for a summer associate I will be looking to take a leave of absence from my current employer to hopefully work with a firm this summer for more firm side experience.

Call me on google voice! or reach me by email at brian at digitalinquest.com