SCILS Initiative in Inter-Disciplinary Study of Issues in Privacy and Security

Paul Kantor (paul.kantor@rutgers.edu) Coordinator

ABSTRACT

In New Jersey at the Rutgers Central Campus, The School of Communication, Library and Information Studies (SCILS) provides a unique strong nexus for the interplay of human behavioral sciences, information processing and retrieval, natural and computer languages, and organizational research. This nexus is relevant to many kinds of research initiatives. Here we set out a pilot plan focusing on issues of homeland security, safety, privacy, and related issues. We hope to expand this into a recognized center of excellence, either as a stand-alone unit, or in collaboration with other units at Rutgers.

We have received $100,000 from Rutgers University to be used for start-up packages, visitors, and workshops. The research integrates Information Science, Human Behavior, and Individual and Organizational Communication and Training. SCILS will generate at least $2M in matching external funding. This program is distinguished in these ways: (1) the human user of systems, and/or the users community is at the focus of research, (2) concern with issues of privacy and individual rights are given primary weight in analyzing alternatives and designing systems, and (3) our track record of collaboration by scholars across multiple disciplines ensures that funds will be used to build new bridges, rather than to fortify existing castles.

NARRATIVE

A major concern in America today is the balance between safety/security, and the cherished rights of personal liberty. With the communication revolution represented by the Internet it is becoming possible for agents, both benign and malevolent, to collaborate with stunning speed and accuracy. The disciplines at SCILS which bear on these issues are concerned both with the technologies of managing information, and with the human and organizational issues on which the impact of information depends.

SCIENTIFIC COMPONENTS

Communication

Current scientific work in Communication includes work based in psychology, which detects and classifies almost imperceptible changes in expression that reveal lying or stress (Frank). This work builds on research by Frank and others in establishing that there are fundamental, biologically controlled, micro-expressions which are reflective of a person’s internal emotional state in a way that transcends cultural and national conditioning. This scientific knowledge has direct application in safety and security settings, and substantial work at SCILS has focused on training individuals to be accurate perceivers of the underlying emotional state, particularly in stressful situations. A second major stream of research in communication deals with persuasive messages and cultural contexts, and addresses the problem of conveying facts and information in a way that is accessible and persuasive to the people most affected. Current work has focused on the creation of effective messages for preventing alcohol abuse, encouraging organ donation, and understanding responses to genetically engineered foods. The principles established in these scientific studies can also be applied to disseminating information about health and safety, emergency preparedness and response plans, and other messages that strengthen the nation's security. (Morgan, Stewart, Lederman, Greene, Yanovitzky, McInerney, Kubey, Steiner).

Information Science

Current research activity in Information science focuses on both the behavioral and the computational issues in the use of networked information resources. Computational algorithms are used, for example, to build automatic models of the qualitative judgments that individuals make about the messages that they receive. This work builds on statistical analysis, and machine learning methods such as Support Vector Machines, Classification Trees, and the standard (parametric) models of regression and analysis of variance. This body of principles is being applied to: detecting patterns and significant events in message streams (Kantor); personalizing the interface for a specific user, based on implicit sources of evidence (Belkin); or explicit sources of evidence (Kantor); and tailoring information systems to serve collections or collaborations of users (Kantor, Zhang). Other projects and programs address understanding streams of information by reference to repositories of prior knowledge, with particular attention to discovering novel events (Lesk) and novel methods for encoding dynamic information (Kantor), all of which are central to online real-time processing of information relevant to safety and security. Additional research on natural language, and automatic interpretation of characteristics of messages complements the machine learning work by providing new insights into the characteristics of messages that provide the basis for human evaluation and judgment (Wacholder). Yet other research projects deal with improving the visualization of complex sets of data (Spoerri) and on formal extensions of XML to improve management of information across the world wide web, particularly when diverse ontologies must be integrated, as occurs in trans-national work, and even in the integration of the diverse domestic databases relevant to homeland security (Muresan).

Organizational Science

Finally, the direct impact of all these scientific advances will be strongly conditioned by educational, organizational and dissemination policies. These issues are the focus of research in the areas of leadership, technology diffusion, and effective education for the user of technology. This research, which has substantial external funding from corporate, foundation and federal sources, considers issues ranging from individual leadership and effectiveness to the effectiveness of small organizational units and large complex organizations. SCILS researchers are recognized nationally, and advise the board for the national quality awards in the non-profit sector (Ruben, Aakhus, Bunz, McInerney, Winston, Wacholder). As organizations become more effective in protecting our collective security, it remains vital, in the American tradition, to provide counterbalancing attention to individualization (Belkin), and to the rights of privacy (Katz).

These specific research activities are currently supported by over $5 million in funding from agencies and foundations. In sum, the scientific principles and research methods used by SCILS faculty and students range from computational/mathematical models through quantitative studies of individual and group behavior, to qualitative investigation of information behavior, and of organizational activities. The enterprise is inherently interdisciplinary, and the SCILS organization is committed to maintaining an open door policy to research collaborators with whatever strengths and perspectives are needed to advance Homeland Security.

THE PROPOSED ACTIVITIES

The work at SCILS will serve as a nucleus for collaboration across academic disciplines and institutions, building on existing working relationships with DIMACS (The Center for Discrete Mathematics and Computer Sciences), the Division of Computer Sciences, the Rutgers Center for Cognitive Science, the Bloustein School of Public Policy, the School of Engineering, the Graduate School of Education, and other departments at Rutgers. Other universities with which we are collaborating include Princeton, Rutgers Newark, SUNY Albany, University of Massachusetts at Amherst, UC San Francisco, University of San Francisco, Salk Institute, University of Pittsburgh, Carnegie-Mellon University, Bond University (Australia), Portsmouth University (UK), SUNY Buffalo, University of Houston, the New York Academy of Medicine, and Washington University. Nationally, we propose to increase Rutgers visibility as focus for visiting scholars and topical workshops. We will reach out to partners in New Jerseys large industrial community. Examples of potential partners include the Internet search engine companies (Lesk, Belkin, Kantor, Zhang), the pharmaceutical industry (Lesk, McInerney) where several pilot studies related to organizational response to technology have been completed, and the communications industry, particularly ATT Bell Labs. Current partners across the country, for the work on deception detection, include Quasar, Jet Propulsion Laboratory, Erica, Inc., Honeywell, Institute of Analytic Interviewing, and Boeing (Frank) and for Information Science, TEOMA/AskJeeves (Belkin).

We are distinguished from other Rutgers programs addressing security in these important ways:

(1) at SCILS, the human user of systems, and/or the users community is at the focus of research

(2) at SCILS, concern with issues of privacy and individual rights are given primary weight in analyzing alternatives and designing systems.

(3) at SCILS, the track record of collaboration by scholars across multiple disciplines ensures that funds will be used to build new bridges, rather than to fortify existing castles.

Participating Faculty at SCILS.

Aakhus, Mark [Organizational Issues]

Belkin, Nicholas [Privacy and Personalization],

Lesk, Michael [Information]

Kantor Paul [Filtering, Interactivity]

Katz James [Privacy]

McInerney, Claire [Persuasive Messages; Workforce Development]

Muresan, Gheorghe [XML extensions]

Ruben, Brent [Organizational Issues]

Spoerri, Anselm [Visualization]

Stewart, Lea [Persuasive Messages]

Wacholder, Nina [Natural Language Issues]

Yanovitzky, Itzhak [Persuasive Messages]

Zhang, Xiangmin [Collaboration]