Paulo Cesar G. Costa

Paulo Costa

Research Interests

My research focus encompasses four synergistic components: 1) the primary research focus on probabilistic knowledge representation and reasoning; 2) the extended research on evaluation of uncertainty in information fusion; 3) the extended research on safety and security of cyber physical systems and 4) the interdisciplinary efforts combining these areas to transition methodological research to practical significance. The latter currently includes establishing the return of investment (ROI) on advanced manufacturing and supply chains, development of ad-hoc intrusion detection systems, safety and security of intelligent transportation systems, cybersecurity experimentation, healthcare automation, and cyber-physical security of smart power grids.

The major area of my research is on the use of Bayesian probabilistic reasoning as a means to achieve better information technology systems, with a special focus on decision support and multi-source data fusion. As a result, I am interested on the diverse areas of knowledge that contribute to this broad challenge (e.g. Data Mining, Utility theory, Markov processes, Game theory, etc), while also flirting with the different domains in which I believe Bayesian reasoning can make a difference (e.g. BioGenetics, C3I systems, Web applications, etc). At present time, I am working on or interested in:

Major research initiatives in which I have some degree of participation

CyManII

The Cybersecurity Manufacturing Innovation Institute (CyManII) is an inclusive national research Institute with major leading research universities in cybersecurity, smart and energy efficient manufacturing, and deep expertise in research and development, supply chains, factory automation, and workforce development. Led by The University of Texas at San Antonio, CyManII leverages the strongest Department of Energy National Laboratories in this area with Oak Ridge National Laboratory leading the nation in advanced manufacturing, Idaho National Laboratory leading in cybersecurity of industrial control systems and physical infrastructure, and Sandia National Laboratory leading the nation in cybersecurity of supply chain management. Funded by the DOE, CyManII aggregates the most advanced research institutions in smart and advanced manufacturing, securing automation and supply chains, workforce development, and cybersecurity. The research team brings to bear the most powerful expertise and infrastructure needed to secure the digital transformation that will continue to propel the U.S. in innovated research in manufacturing for decades.

ETURWG

The Evaluation of Technologies for Uncertainty Reasoning Working Group (ETURWG) is an official working group of the International Society for Information Fusion (ISIF), which was formalized in July 2011. The group provides a forum for its members to collectively address a common need for the ISIF community, coordinate with researchers in the area, and evaluate techniques for assessing, managing, and reducing uncertainty. The main aspects involved are:

The goal of the ETURWG is to bring together experts, researchers, and practitioners from the Fusion community to leverage the advances and developments in the area of evaluation of uncertainty representation to address the problem of evaluating uncertainty representation and reasoning approaches for High Level Information Fusion (HLIF) systems. The group will provide a forum for its members to collectively address a common need for the ISIF community, coordinate with researchers in the area, and and evaluate techniques for assessing, managing, and reducing uncertainty.

Recombinant AI: Exploiting Heterogeneous Data Fusion with Ontological Frameworks and NLP

This project focused on developing analytical frameworks (e.g., ontological framework) to enable heterogenous data exploitation and fusion in support of Recombinant AI objectives. It builds on the outcomes of Natural Language Processing (NLP) applied to various documents in a given domain (e.g., data extraction, indexing, and translation etc.), an ontology framework, for example, will identify key entities of interest in that domain and how these entities are interlinked towards inferring root causes or potential future courses of action. It set foundations for exploiting and fusing heterogeneous data to overcome limitations of missing data and/or extract new information from disparate and siloed data sets.

ADS-Bsec

Automatic Dependent Surveillance Broadcast (ADS-B) is one the most prominent air traffic surveillance technologies that emerged in recent years. It can be used not only as a complement to existing radar-based technologies to locate aircraft, but also as a replacement in areas without radar coverage. However, a major flaw in the deployment of this technology is its reliance on clear-text, open broadcast of messages, which makes it vulnerable to attackers. ADS-Bsec is a multidisciplinary research effort led by Dr. Duminda Wijesekera and myself that resulted in a holistic framework that solves the root cause of ADS-B problems by enforcing the much-needed authenticity and integrity in ADS-B messages. It does so without changing the message format, and along with a proper key exchange mechanism. The framework also leverages results from radio-location and risk-adaptive techniques in order to prevent attacks targeting ADS-B. An important characteristic of ADS-Bsec is that it was mostly implemented using a test bed with actual prototype radios located at the GMU Radar and Radio Engineering Lab (RARE Lab). As a result, the test environment mimics possible real flight conditions with enough realism for evaluating cyber-physical characteristics commonly used in attacks, some of which cannot be captured in purely soft simulation testbets. The framework is a promising approach for integrating ADS-B to NextGen infrastructure in a secure and affordable fashion, and its effort resulted in a PhD Dissertation (Dr. Thabet Kacem) and a US Patent Issued. This area of research, as highlighted earlier in this document, is currently being explored in the AFSOR-funded project Secured and Robust Communications on Urban Air Mobility Networks (SRCUMAN).

PROGNOS

PROGNOS is the abbreviation of PRobabilistic Ontology for Naval Operations Systems. The project was supported by the Office of Naval Research (ONR) via its thrust area: Automated Information Integration (sub-area: Predictive Situation Assessment to Provide Impact Assessment). GMU team used probabilistic ontology to model various environment and situation, and is developing mathematically rigorous, computationally efficient algorithms, and tools with GUIs for predictive analysis, and situation assessment.

PR-OWL

One of the major limitations within the development efforts of the Semantic Web is the lack of a comprehensive, principled formalism for representing and reasoning under uncertainty. That is, current SW research is focused on classical logic and some of its variants, failing to provide a means to deal with the many forms in which knowledge cannot be discharged or ignored (e.g. incomplete, ambiguous, or dissonant evidence).

PR-OWL (Probabilistic OWL) is one of the approaches aimed to solve the above limitation using Bayesian methods for representing and reasoning under uncertain knowledge. The framework of PR-OWL is Multi-Entity Bayesian Networks (MEBN), a First-Order Bayesian Logic that provides the mathematical basis for merging the expressiveness of Classical First-Order Logic with the power and flexibility of Bayesian Probability.

Information on both PR-OWL and MEBN can be found on the PR-OWL website.

UnBBayes-MEBN

UnBBayes is a probabilistic network framework written in Java. It's composed by a inference engine, a GUI editor, an API, and a learning environment. The algorithms used are based on strong junction tree method and measure and search (K2 & B).

Currently I am working with the UnBBayes folks at the University of Brasilia in the development of a GUI and a reasoner for opening, creating, saving, and reasoning with probabilistic ontologies in PR-OWL format. Files saved in UnBBayes-MEBN are backward compatible with "legacy" (non-probabilistic) ontologies so it can be opened in ontology editors such as Protègè or Swoop.

URSW

The focus of the Uncertainty Reasoning for the Semantic Web workshop (URSW) is to address the challenge of representing and reasoning with incomplete, uncertain information in the context of Semantic Web development. The workshop is held at the International Semantic Web Conference (ISWC).

ISWC is a major international forum for presenting visionary research on all aspects of th Semantic Web. The Uncertainty Reasoning Workshop is an exciting opportunity fo collaboration and cross-fertilization between the uncertainty reasoning community and the Semantic Web community

The first workshop (URSW 2005) was held on the fourth ISWC, in Galway, Ireland. The second workshop (URSW 2006) was held at the fifth ISWC, in Athens, GA, USA. The third workshop (URSW2007) will be held at the sixth ISWC, in Busan, Korea.

URW3-XG Incubator Group

I am working as an invited expert in the W3C's Uncertainty Reasoning for the World Wide Web (URW3) Incubator Group, which is part of the Incubator Activity. The group's mission is to better define the challenge of reasoning with and representing uncertain information available through the World Wide Web and related WWW technologies. More information can be found in the URW3-XG Charter.

Academic Service

Current

Past

Contact Me | ©2005-2014 Paulo C. G. Costa | Last Updated on April 4, 2014.

Valid XHTML 1.0!