You are here

Aris Ouksel

Head and shoulders photo of Aris Ouksel
  • Professor of Information and Decision Sciences

Office: 2411 UH
Phone: (312) 996-0771
 

Education

  • PhD, Northwestern University, 1985
  • MsC, Northwestern University

Experience

  • Director of the Management Information Systems PhD program at the CBA
  • Director of the Center for Management of Information Technology and Telecommunications at the CBA

Research & Publications

Research Interests
  • Information technology strategic planning,
  • Business process reengineering,
  • Organizational learning and performance,
  • Information technology diffusion,
  • Information economics
  • Price discrimination and effect on markets and information
  • Non-Disclosure and impacts on Automated Exchanges
  • Mobile ad-hoc networks
  • wireless and mobile sensor networks and data management
  • Peer-to-peer data management
  • Semantic issues on the web and in virtual inter-organizational information systems
Publications/Projects

1.  Information Economics

1.1  Business Intelligence and Its Impacts on Price Discrimination and Social Welfare

Business intelligence tools have enabled novel and relatively low cost capabilities to collect and analyze vast amount of customer information. Accumulation of customer specific information along with transactional data empowers firms to categorize customers into segments and offer customized prices. Our project studies the impact of price discrimination and market segmentation on competition and consumer purchase behavior in a game theoretic model with two asymmetric firms. Recent results show that at equilibrium, both firms price discriminate and segment the market. Contrary to previous price discrimination and market segmentation findings, the game is not necessarily a prisoner's dilemma. The firm dominating the industry is likely to improve its profits at the expense of the rival firm, and consumer welfare will increase with segmentation. We define two fundamental parameters, market dominance and the technology cost to industry dominance ratio, to drive segmentation technology adoption decisions, as a basis for our analytical approach. There are a number of issues remaining to investigate. In particular, we investigate the effect of business intelligence quality (incompleteness and uncertainty) in correctly segmenting the population and business intelligence technology investments on competitiveness under different time horizons. Other investigations examine the impact of cost differentiation on market structure, assuming price discrimination and market segmentation. This problem has not at all been examined in the literature. Recent results show that costs matter, and have a significant on market structure.

1.2  Smart Pricing and Market Formation in Hybrid Cellular Networks

Demand  for wireless communication services is growing faster than network providers can add capacity. At peak demand, networks must limit service consumption; networks that can support more customers while maintaining quality of service will have an important advantage over their competitors. As a partial solution to this capacity problem, we propose incentivizing cellular network customers to join smaller-scale peer-to-peer WiFi networks. We present a network architecture built around three components: a voluntary participation system that allows switching between the cellular and peer-to-peer networks; our Self-Balancing Supply/Demand protocol for information search and discovery; and a cap-and-trade model for minimizing the total incentives required to support the operation of the peer-to-peer network. 

1.3  “Cap-and-Trade and Congestion Pricing for Smart Vehicles and other applications,” UIC-Funded projects (PI). Winner of the 2011 US Department of Transportation Challenge on Connected Vehicle Technology.

A market-driven cap-and-trade policy is considered the most economically efficient approach to limiting emission of pollutants. The policy permits those who can reduce emissions at the least cost to sell their pollutant credits to those facing higher marginal costs, and thereby achieves economic efficiencies – lowest cost and least distortion. While economists concede the efficiency of such a cap-and-trade policy, its administration and implementation presents significant challenges, including competitive dynamic resource management, real-time auditing, behavior forecasting, and behavior incentive structures. Mobile environments further exacerbate the complexity of these challenges. In this proposal, we investigate the design of a mobile ad-hoc network-based novel system that combines innovative computing, economics, and networking approaches for on-the-fly competitive cap-and-trade pollution in vehicular networks to support sustainable transportation policies. In particular, we develop game-theoretic-based coordination mechanisms for realizing an efficient cap-and-trade dynamic resource allocation system, which exploits traffic, congestion, and ride-sharing parameters. The outcome will be a dynamic system that monitors pollution and allows drivers and owners to competitively negotiate their pollution quotas in real time with the vehicles near and around them, and with the fixed sensing infrastructure on the roads, by forming a wireless ad-hoc network. This vehicular network represents a futuristic pure peer-to-peer mobile environment. The outcomes however, apply to intermediate architectures where the components are only partially mobile, and to many other applications such as energy consumption, bandwidth/computing-power/services management in mobile infotainment and cloud computing. 

2. BIG Data Analytics in Cybersecurity 

2-1 . Identification of Emerging Threats in Text Messaging, PUF (Partner University Fund), funded by the French Government in partnership with University of Arizona and University of Lyon, France.
 
Relevant Themes: community, context, social networking, risk management
 
The field of information security has been challenged by an ever-increasing rate of threats. Studies have shown that the rate of emergence of new threats is on the rise, to such an extent that information security organizations continuously struggle to develop ways to identify and protect against such threats. The main reason for this lies in the traditional method of threat identification, which is designed to identify known threats. This typically involves compiling extensive lists to identify all known possibilities. Each is weighed by the potential risk that each threat poses to the organization or their assets, especially those that have a high probability of occurring. Once this is done, organizations identify the threat signatures they need to look for to identify the likely known threats. Unfortunately, this process can be expensive, complex and ineffective in identifying new or invisible threats.
A threat is a hazardous condition that can take advantage of vulnerabilities (or weaknesses) in what is referred to as an object (e.g. an asset, a system, an operation, a person, etc.). Threats emerge through what is known as an agent (anauthor, idea, or actor) who can take the form of a person, system, application or some other entity. Threats are often recognized through a known signature – a trademark of properties that characterize their existence. Once their presence is identified, the potential of an attack exists.

The problem of identifying new threat signatures is one that has tested the security resources of both government agencies and private enterprises alike. The (http://liris.cnrs.fr/cyber/at) d which threats evolve has outpaced security protection mechanisms, and to this day continues to threaten not only enterprise operations, but also the entire critical infrastructure of the U.S. The inability to tackle this problem has been clearly acknowledged as a looming cyber-threat [3].  It has also been recognized that in homeland security, no agency has now or will ever have, all the needed information or responsibility to combat terrorist threats [4]. Traditionally, even such well known computer network threats such as Distributed Denial of Service (DDoS) attacks, malware, spoofing and botnets, to name a few, can emerge with new types of signatures that can make identification quite difficult.

Recently, the Defense Advanced Research Projects Agency (DARPA) has recognized the potential for leveraging social networking capabilities to address security problems [5]. Both state local and governments have recognized effective peer-to-peer text communication and information disseminated through social interaction as a powerful security tool. However, leveraging the capabilities of social media technology for threat analysis has yet to reach its full potential [4]. Identifying a new threat emergence from text information generated across covert groups such as terrorists has posed challenges to data fusion and semantic analyses.  What is required is a structured analytical framework of methods and tools that can take text communication from seemingly unrelated agents and dynamically assess the potential of new threat signatures, along with their associated meaning and contexts. Such a framework can arm security organizations with enhanced filtering methods to identify likely threats emerging from covert and corporate groups, with greater levels of assurance.

2-2  “Resource Management amongst Autonomous Command Centers: Application to Crisis Management,” NSF-SGER-0743331.
 
Mobile ad-hoc networks (MANETs) promise to change the distributed systems communication paradigm and the ways people access and manipulate information, as well as application development. Proposals abound to deploy them in many applications, including, homeland security, disaster recovery, transportation, and military battlefield management, to cite a few. However, there is now widespread recognition that their viability in many applications is dependent on providing robust solutions to key core issues of these systems including: (i) reliable and efficient information dissemination and discovery protocols; (ii) transaction and coordination management mechanisms for complex applications; (iii) trust-aware routing and synchronization protocols to support reliable and secure communication; and (iv) flexibility to adapt to different organizational forms.

3 Big Data/Analytics in Wireless Sensor Networks and Mobile Networks with applications to: 
disaster management, health, transportation, and supply-chain management

Automated Information Factories
 
There has been a growing trend toward the automated generation of massive data at multiple distributed locations. Examples include systems to monitor the physical world, such as wireless sensor networks, and systems to monitor complex infrastructures, such as distributed Internet monitors.  This trend will likely continue, leading to a future of computing that is data-rich, heterogeneous, distributed, and rife with uncertainty. 

Most information available today on the Internet is fabricated by the people data entry – this includes the text in web pages, the sound and video clips on various popular media-sharing sites, and typical corporate media.  Increasingly, though, the future is pointing toward automatic information generators such as sensor networks, traffic monitoring systems, email spam generators and mutators, network firewall logs, and feature-extraction systems that can annotate media streaming from inexpensive unattended monitoring devices. The volumes of information produced by these systems will quickly dwarf anything produced collectively by humans. While traditional media will continue to be fabricated, it will be only a small fraction of the volume of information generated.

Our project will address the challenges posed by the deployment and use of networks of fixed and mobile heterogeneous smart sensing heterogeneous devices (writ large as these may involve cameras and other sophisticated devices) to monitor geographic areas containing possibly moving objects (themselves part of and responsive to the environment). These sensors capture information on a plethora of composite spatio-temporal events of interest, which when fused and analyzed in a specific application, may trigger the execution of pre-defined continuous queries and responsive actions. UNCERTAINTY: Current technology cannot deliver the responsiveness demanded by the envisioned monitoring and surveillance systems that require activity orchestration to support real-world application-related decisions and/or triggering of context-driven query generation for refining the quality of information about the monitored phenomenon. An effective overall system must factor in competing demands on sensor network resources and their best allocation to sensing the environment and analyzing the sensed data. Current state of the art in sensor networking technologies, query processing, and information fusion, falls short in facilitating the management and integration of such evolving heterogeneous information in WSNs. To meet the multiple research challenges raised by this problem in networking, data management, control, and signal analysis, it is imperative to develop an integrated approach to realize effective sensor net applications. Applications include for example, in urban settings, the sensors monitoring various structural parameters of bridges should trigger frequent sampling in cases where heavier traffic load is predicted in the near-future.  Other applications include vehicular networks, eco-settings, and ubiquitous health monitoring.

3-2  IGERT: Graduate Program in Computational Transportation Science. NSF-DGE-0549489, 2006-12.
 
This project is a Ph.D. Research program to train research specialists in the Information Technology aspects of Transportation Science. The IGERT fellows will develop technologies in which sensors, travelers’ computers (e.g., PDAs), in-vehicle computers, and computers in the static infrastructure are integrated into a collaborative environment. We are currently investigating how these technologies are adopted, and the implications of such adoption. The envisioned environment will enable solutions to transportation problems ranging from dynamic ride-sharing, real-time multi-modal routing and navigation, to autonomous/assisted driving, to inferring travel patterns via data mining. Basic research in information management, communication, software architecture, economic modeling tools, human factors, traffic prediction, and transportation planning and policy is essential for founding a new discipline that will integrate millions of disparate, high mobility computers and sensors into a collaborative system. Presently, the web performs a similar integration function for stationary computers.

3-3 . “Context-Aware Computing with Application to Public Health Management”, NSF-IIS-0326884, 2003-2009. 
 
This project deals with opportunistic discovery and dissemination of resource-information as well as incentive mechanisms to support increased cooperation among the vehicles in ad-hoc vehicle-to-vehicle mobile networks. Such networks will be supported by short-range wireless communication capabilities such as 802.11, Ultra-Wide-Band (UWB), and Bluetooth. Our work in this area has been well received by researchers in this area, as illustrated by the best paper award at an important international conference. We have extended the Publish/Subscribe paradigm to support a self-organizing approach to content dissemination and discovery in mobile environments. The approach is based on the economic supply/demand model to regulate the dissemination of services and requests over the networks. Applications include health-care delivery and public health management, command and control centers in battlefield management, and transportation science.
 
4. “Resource Management amongst Autonomous Command Centers: Application to Crisis Management,” NSF-SGER-0743331.
 
Mobile ad-hoc networks (MANETs) promise to change the distributed systems communication paradigm and the ways people access and manipulate information, as well as application development. Proposals abound to deploy them in many applications, including, homeland security, disaster recovery, transportation, and military battlefield management, to cite a few. However, there is now widespread recognition that their viability in many applications is dependent on providing robust solutions to key core issues of these systems including: (i) reliable and efficient information dissemination and discovery protocols; (ii) transaction and coordination management mechanisms for complex applications; (iii) trust-aware routing and synchronization protocols to support reliable and secure communication; and (iv) flexibility to adapt to different organizational forms.