The World Wide Web (WWW) and more generally the Internet are providing new ways of communicating and of doing business. There are many facets to the developments and many areas in which research and teaching related to these developments might be carried out. The School of Information Sciences at the University of Pittsburgh is one of the places where these new developments should be explored. Like SIS, the Graduate School of Business and the College of Business Administration have a number of reasons to be interested in Electronic Commerce. Preliminary versions of this paper have been shared with a number of our colleagues at KGSB and these same individuals are working together on a number of fronts. (Dr. Michael Spring has included faculty from Katz and from Computer Science in a project recently funded by Hewlett Packard on the development of E-Services, which many in the field consider to be the infrastructural basis of the next generation of E-Commerce.)
This proposal is restricted in looking at E-Commerce issues from the information systems and technology focus of the Department of Information Science and Telecommunications. We are seeking support for this initiative from the Provost's Office because the Department's resources have been stretched exceedingly thin by rapidly growing enrollments and large research grants. The faculty and administration have reviewed an internal version of this proposal and endorsed it. We believe that an electronic commerce initiative can be mounted at a reasonable cost with a significant payoff to the department, school, and university.
To situate that work, we begin with a definition of commerce:
1: social intercourse: interchange of ideas, opinions, and sentiments
2: the exchange or buying and selling of commodities on a large scale involving
transportation from place to place.
Basically,
commerce includes the interchange of ideas and information and the interchange
of goods. A strong case could be made
for the fact that in an increasing number of instances, the goods exchanged are
ideas. In these cases, definitions one
and two are merged. Information Science
was created to bring together the disciplines necessary to address the
development of e-commerce, both in the narrow and broad definitions of the
terms. We as a department have already
made several forays into the fundamental theory that underlies any study of
e-commerce. Standards are fundamental
to E-commerce and we offered the first course in the country devoted to IT
standards at the graduate level.
Documents are central to E-commerce and we have offered a course on
Document Processing for more than 15 years now. E-commerce systems are interactive client server systems; we
offer courses at both the graduate and undergraduate level in Interactive
Systems and at the graduate level in Client Server Systems. Databases are the heart of E-commerce and we
offer multiple courses related to relational, geographic, object oriented, and
distributed databases.
Set out
below are a list of the basic requirements for four programs in this area. Also
listed are some rough course descriptions and a series of the research
questions that we think might contribute to developments in this area.
The
curriculum might be conceived of as operating at four levels:
· general continuing education/certificate program ( C )
· focus of an undergraduate degree program ( U )
· track or specialization in an MSIS program ( G )
· specialized executive masters program for company sponsored individuals ( X )
Rather than
defining each of these curricula in detail, a map of prerequisites courses
along with specialized courses are set out below. This is a rough first cut outline and will need to be fleshed out
and linked as the courses are detailed.
For the specialized courses to be developed (or evolved from existing
courses in the curriculum), a brief description is provided suggesting the
kinds of topics that would be covered.
The courses are keyed back to the programs in which they might be
offered by the letters in parentheses after the course title. In each case, the details and specific
content would be tailored to the level at which they are offered. We are obviously not suggesting that the
same course be offered at different levels. (It is important to keep in mind
that the limited time used to develop this proposal did not allow us to include
course from economic, business, or industrial engineering perspectives. The courses listed here only reflect the
information science perspectives.)
The reader
will notice that there is a strong emphasis on security in this program. Several aspects of system and network
security are critical to e-commerce.
Commercial sites must be hardened to prevent various forms of
intrusion. Transactions and data stores
must assure an individual's right to privacy.
E-commerce demands authentication, encryption of critical data,
assurance of data integrity and methods for copyright protection and assurance
of information integrity. At the same
time, security in network based applications is broader and more complex. We have had discussions over the last
several years with CERT and SEI about the development of some focused effort in
security. One important spin-off of the
E-commerce effort might well be an additional program focused on security.
|
Course and Tracks for E-commerce: A DIST Perspective |
|||||
|
|
Certificate |
Undergraduate |
Masters |
Exec. Masters |
|
|
Prerequisite Courses |
|||||
|
Data Structures |
|
X |
|
|
|
|
Programming Design and
Software Tools |
|
X |
|
|
|
|
Telecommunications |
|
X |
|
|
|
|
Systems Analysis and
Design |
|
|
X |
|
|
|
Database Management |
|
|
X |
X |
|
|
Interactive System
Analysis and Design |
|
|
X |
|
|
|
Fundamentals of
Telecommunications |
|
|
X |
X |
|
|
Client Server Systems |
|
|
X |
|
|
|
Special Focus Courses |
|||||
|
Web Authoring |
X |
X |
|
X |
|
|
Web Languages |
X |
X |
X |
X |
|
|
Internet Std. and
Protocols |
|
|
X |
X |
|
|
Web EC Security |
|
X |
X |
X |
|
|
Security in Network
Environments. |
|
|
X |
|
|
|
Advanced Crypt.,
Biometrics, & Encoding |
|
|
X |
|
|
|
Building E-commerce
Sites |
X |
X |
X |
X |
|
|
Web Databases |
X |
|
X |
X |
|
|
Collaboration in a
Network Environ |
|
X |
X |
|
|
|
Distributed Object
Computing |
|
|
X |
|
|
|
WWW/Internet based
Information Systems |
X |
|
X |
X |
|
Web Authoring (C-U-X)
This course
would address the issues pertaining to the analysis, design, and management of
websites. It would introduce web
authoring languages and web authoring tools.
It would cover simple active pages -- web scripted pages, and simple
server applications -- CGI programs. It
would cover HTML as a language and XML as a system of languages. Cascading Style Sheets and the XML link
language would also be introduced.
Web Languages(C-U-X)
This course
would introduce languages that have been optimized for the creation of
environments and activities on the WWW.
It would cover in detail the structure and theory of XML, in particular
by contrast with the parent language SGML.
The evolution of languages for structured document interchange including
interscript, postscript, EPSF, and PDF, ODA, and others will be reviewed. The development of the Resource Description
Framework and the denigration of EDI standards for data interchange will be
explored as well. Finally the standards
for Design Syntax Specification Semantics Language and the Hytime linking
standards will be explored both in their broad theoretical structure and in
terms of the manifestation as subordinate web standards
Internet Standards and Protocols (G-X)
The internet
protocols that support electronic commerce generally will be explored. These will include, but not be limited to
addressing standards (IP), transport standards (TCP, ATM), applications
standards (SMTP, SNMP, HTTP, FTP, TELNET, etc.). In addition, this course will address the emergence of
distributed object standards such as COM and DCOM, CORBA, and RMI. The development of infrastructure protocols
such as LDAP, DES, and Digital Certificate standards will also be
addressed. The course will provide
students with the opportunity to explore these standards in the course of real
application development.
Web E-Commerce Security Issues (U-G-X)
This is a
narrowly focused course on Web security issues pertaining to e-commerce that
will focus on server side vulnerabilities, servlet/CGI precautions, general
applications of secure communications including the use of passwords, cookies,
secure socket layer communication, and certificates. The course will provide the student with hands on applications of
these technologies with an eye to developing the set of skills required to do
competent management of a corporate website.
Security in a Networked Environment (G)
This course
will address the general issue of security in a networked environment. The course will broadly be broken into three
parts. The first section of the course
will identify the basic frameworks for analyzing systems security. This includes treatment of both threat
taxonomies and defensive frameworks.
The second section of the course will address in greater detail the
nature and signature of key threats including but not limited to: corporate
espionage, hacker pranks, denial of service attacks, internal attacks, fraud,
etc. The final section of the course
will address various approaches to risk management including the development or robust applications, the
use of packet filtering and proxy dependent firewalls, the use of scanners and
other tools to find weakness and the use of traffic analysis and other
techniques to identify intrusions.
Advanced Cryptology, Biometrics, Data encoding, and Data transfer (G)
This course
will address issues of cryptology including the history of cryptology -- the
Ceasar code, and the seminal papers on cryptology, e.g. Merkel and
Helman(1978), Shamir(1984), Shannon(1949), Diffie and Helman(1977), and Rivest,
Shamir, and and Adelman(1978). The
course will explore block and substitution methods, single and dual key
cryptography, public and private key cryptography, etc. Biometrics will include a review and development
in the area of authentication including retnal scans, finger and palm prints,
voice pattern recognition, handwriting recognition, etc. Data encoding and data transfer developments
including both media types and compression algorithms, watermarks and other
verification techniques, and various approaches to insuring data integrity and
interpretability over time. Beyond RDF,
ASN.1 and other generalized approaches, the course will also address semantic
forms of specialized encoding such as the Dublin core, TEI, and PICS.
Building E-Commerce Websites (C-U-G-X)
The course
will address the important issues in building a website for e-commerce. It will cover the various server side
technologies available to connect partners and consumers to corporate information
stores. It will generally be broken
down into three sections, end user e-commerce, business to business e-commerce,
and enterprise computing -- e.g. supply chain models. The course will address how web managers for medium to large size
organizations should monitor network traffic for system violations, develop
policy for internal employee use of the web.
It will also address how to link order data and credit information in a
secure way. The course will also
address problems of data security and integrity for organizations engaged in
enterprise wide computing applications.
Web Databases(G-X)
This course
will treat Web-Database interfaces, especially tools and languages for making
data stored in relational databases (such as product catalogs) available on the
Web. It will also cover Web server performance, and the precomputation or
caching of database-generated web pages. It will then address web-site design
techniques and Web-Site Management Systems, which are needed for organizing web
sites consisting of hundreds or thousands of pages. In addition, issues of
customization of web pages and data mining for web databases will be discussed.
Finally, XML databases, data extraction from web pages, and the integration of
heterogeneous information sources will be explored.
Collaboration in a Network Environment (U-G)
This course
will address a variety of topics in the area of collaboration as it related to
network environments. The course will
address the notion of organizational capital -- physical, financial,
intellectual, social, and knowledge.
How research in each of these areas is moving forward to support
collaboration will be addressed.
Interactive system design, embedded computing, data capture methods,
analog to digital data conversion data analysis and data mining will all be
reviewed. The development of
visualizations, visual languages, virtual worlds, social environments, etc.
will be reviewed in terms of how they support collaboration in these areas. Key applications of network based collaboration
tools will be presented and explored including conferencing tools, authoring
tools, decision making tools, brainstorming tools, help desk environments,
instructional tools, etc.
Distributed Object Computing (G)
This course
will look at advanced techniques to client server computing. New algorithms and techniques will be
explored. CORBA, RMI, and DCOM will be
addressed in detail. Reflection and
other interface query mechanisms will be reviewed. Operating system changes required for distributed systems will be
reviewed
WWW/Internet based Information Systems (C-G-X)
This course
will explore internet and intranet technologies. Analysis and Design techniques will focus on predicting new and
emerging markets, issues pertaining to global business execution. On the intranet side, the course will
address workflow automation, reengineering, and cooperative processing. In
terms of internet businesses, the course will explore wallets and the cost of
efficient transactions, streaming media applications, auctions and bid based
web applications. Finally the course
will explore public service and government issues. These would include the use of the web to provide easy access to
caregiver/family needed information about medical issues, the use of public for
a on the web for the discussion of political issues, reduction of the
information float in public opinion generation, etc.
The
development of a new focus on e-commerce will help faculty to develop or
enhance their research programs. The
following list of research topics is indicative of the kinds of research
initiatives that might be encouraged by increased focus on E-Commerce. The first two topics have been fleshed out
as a set of interrelated questions.
These are followed by a list of additional topics that provide some
indication of the richness of this area.
1- Auctions/negotiations/flexible pricing schemes. This topical area has both theoretical
issues (like options modeling) and practical/behavioral issues. Even at this
late date there is no data on human machine bargaining/negotiation.. Al Roth
(who we lost to Harvard last year) was doing human-human bargaining but the
other issues of human-agent, delegated agent, etc. etc. schemes for determining
price or other attributes of transactions is completely open as a research
area. In the meantime the most
conspicuous local success since Lycos/Wisewire is Openmarkets. This whole area (at least for the next 6
months) is really hot both commercially and for relevant research.
2- Multi-attribute matching, customization, & product
differentiation. As web
increases the efficiency of markets the point is reached where all sales of
commodities accrue to a single seller, margins are eliminated, etc. No one can continue to compete on price alone
when shopbots can effortlessly determine the lowest seller. Therefore vendors must stress unique
products, customization, and features which "break" one attribute
comparisons. Research challenge is to
develop more complex, multi-attribute, approximate matching methods for
identifying & choosing among buyers/sellers. This doesn't need to be zero
sum since utilities for attributes may vary from party to party.
Other topics
· How will approximate matching using XML and other meta-tagging schemes impact information retrieval and data interchange? What are the provenance needs for archival information? How will information retrieval research influence meta data and the development of semantic URL's?
· How will the Web and ubiquitous computing interact?
· How will user profiles be used in internet environments? What is the ultimate form of cookies and wallets
· How will vr/vrml modes of interaction impact purchase decisions in e-commerce. Will the patterns be different for furniture, clothing, etc.
· How can multicast and broadcast applications best be supported in an internet architecture?
· What are the kinds of social capital that might be supported electronically and how?
· How can privacy be guaranteed in an e-commerce environment?
· What is the optimal mix of human and machine interaction in various collaborative environments?
· How may transactional data be used by agents to support individual activity in a network environment?
· How can interface agents learn users needs?
· What are the best query languages, query optimization and storage structures for XML data
· What kinds of tools are needed for sharing knowledge about the web
· What kinds of web query languages, web search agents are needed
· What kinds of navigational tools are needed for the web? What kinds of data is needed to fuel these navigational tools?
· Transaction mechanisms for the web
· How can we design and management of very large web sites?
· What are the key user-interface design issues for web sites?
· Will web-based teaching work? What mechanisms are needed for selling knowledge on the web
· What navigational tools are easiest for e-commerce users?
· How can e-commerce users be profiled?
A full
blown program in E-Commerce will require significant curriculum
development. It will also require
coordination with our colleagues in the other centers. Finally, it will require the consent and
commitment of the senior administration.
Given the unanimous support of the faculty in the department as well as
the interest of faculty in other units, there would appear to be sufficient
support to move forward with this proposal.
We propose
to begin to flesh out the curriculum with the faculty already in place. There is little doubt that there is a
tremendous demand for courses at all the levels defined. Enrollments in our undergraduate and
graduate courses along with the new programs at CMU, PSU and other places are
sufficient evidence of this. The
continued high demand for continuing education programs with costs of 2000/week
of instruction being common provide support for the demand for certificate and
executive masters types of programs.
The cost of
exploring this effort would be on the order of release time for one faculty for
one year with one or two doctoral students supporting the effort to define the
curriculum in more detail and mount one or two of the easy courses. Whether this is one person full time for a
year or four people half time for a half year would not seem to matter. This would allow sufficient energy to
negotiate course outlines and structure curricular changes. In addition, the project would result in a
set of plans for new degree and certificate initiatives.
Based on an
assessment of the curriculum defined, we might choose to simply incorporate
selected courses into the existing curriculum and consider the project
closed. However, it is our hope that
the curriculum, after review, will be deemed appropriate as a target of
opportunity for further development and support. It is anticipated that a plan for a new certificate or degree
program might require three to five new faculty lines. There would seem to be little doubt that
funding of could be made self sufficient almost immediately without any loss of
enrollments in the existing programs.
The long term costs of new faculty lines would be offset by tuition and
research revenues. The short-term costs
of defining the curriculum might be supplemented by state or federal
support. Specifically, we note that the
Governor's Office has allocated Link to Learn funds specifically for the
purpose of developing new curricula in this area. If we were to make an effort to attract funds form this source,
some immediate commitment would need to be made to this concept.