Imagine being the CEO of Barnes and Noble and picking up the Wall Street Journal in July 1995. Grabbing the phone and calling his CIO, the CEO demands to know, "Who is this Bezos?" Today we know, of course, that it was Jeff Bezos, the CEO and Founder of Amazom.com books, "earth's biggest bookstore." The amazing story of Bezos' virtual bookstore teaches many lessons from the business battlefront of today. Out of nowhere and totally unheard of in the book industry, this digital bookstore turned an industry up-side-down, using blitzkrieg tactics. What happened here was more than creating a web site. An intelligent and ubiquitous digital business was conceived and implemented globally. Its system is its business, its business is its system.
The company represents a competitive form not seen before on the business landscape and not possible prior to the ubiquitous, interactive communication network, the World Wide Web. Shattering traditional value chains in the book-selling industry, Amazon.com opened thousands of virtual bookstores in its first months of operation. These virtual bookstores are independently operated by people and organizations (Amazon calls them Associates) that create web sites in their specialty fields and seamlessly interoperate with Amazon.com using the Web as the backbone.
The many tasks needed to establish a virtual bookstore are handled digitally. Once established, intelligent software agents (dubbed Eyes and Ears) notify the virtual bookstore operators and individual customers of new books relevant to their interest. Thousands of activity and sales reports are distributed across the Internet to Associates each week without a drop of ink. Meanwhile, digital sales clerks gently suggest to customers browsing a particular book that others who have bought the book also bought these other titles - a little reference selling. Amazon's high-tech site is also high-touch. It delivers more information than the shopper of a real bookstore can find as real shoppers are limited to what's on the limited shelf-space and reviews and comments simply don't exist while perusing a shelf of books. True to the notion of the virtual corporation, Amazon.com doesn't need physical book stores or warehouses - why, they don't even need books! They compete with information and knowledge-based information systems and outsource the rest of the business in the industry's value chain. While many companies are contemplating the future of electronic business, Amazon.com does it now and makes it seems easy. But if it's so easy, why isn't every business just doing it?. Proper recognition of the true nature of the challenge is the foremost obstacle.
Today's forward thinking CEO recognizes the challenge of eCommerce as a strategic business issue, not just one more technical issue to be delegated to the IS department, perhaps the existing EDI group. Although a company may have reengineered its internal business processes and perhaps painfully installed an ERP system to bring efficiencies to the back office, eCommerce is about reengineering outward-facing processes - industry process reengineering versus business process reengineering (IPR vs BRP), redefining industry boundaries, inventing new industries, repositioning, disintermediation (cannibalizing supply chains), and reintermediation (establishing portals to the Web). It's about 1-to-1 marketing, segmenting a single customer instead of segmenting mass markets. Its about outsourcing customer service to the customer himself. It's about relationship marketing and building communities. It's about honoring the customer not as king, but as dictator who is but one mouse click away from turning his back on you. This is the stuff with which CEOs, not programmers, must grapple.
The transition to the emerging digital economy will not happen over night, and there will be many skirmishes, misfires, and changes of direction along the way. Surprise attacks will shake industries, ala Amazon.com's "hello Barnes and Noble," and E-Trade's "hello Charles Schwab." Pioneer Electronics did something in May 1998 it had never done before and began selling directly to consumers. Not wanting to alienate it's network of 1,200 dealers and 15,000 stores it left out those products they carry, at least for now. Why even risk alienating them? Pioneer had to in response to competitors. Meanwhile GM is bringing to the Web its response to Auto-by-Tel and Carpoint with consumers going online to configure and price its vehicles without bypassing dealers. Instead the site will direct customers to the nearest dealer and let the haggling begin. Stories such as these appear with greater frequency as we step across the threshold to the digital economy.
IT departments have a new customer, the marketing department. While traditional IT has been inward facing, handling the affairs of the back office, successful marketing requires outward, customer-facing, real-time, "killer applications." Marketing itself has witnessed a shift in focus from the product to the customer. No longer is branding of a product the key to success, customers want products made just for their specific needs. As we shift from mass production to mass customization, marketers must know the needs of their individual customers, not the needs of market segments. Remember the Sears catalog? Sears had information about addresses where it could mail its catalog, but, due to the technology of the time, not the intimate details of its customers' buying behavior. The new capability to harness information technology to drill down below the demographics directly to the behavior of individual customers has far reaching implications for companies that don't want to go the way of the Sears catalog. As marketing guru Regis McKenna explains: "It's about giving customers what they want when, where, and how they want it."
The business processes and workflows of commerce are today essentially human phenomena. People are adaptive problem solvers and can deal with new situations "on their feet" and handle incomplete, inconsistent information in real-time, usually making good business decisions based on judgment gained from experience. This is precisely what computers must be empowered to do in eCommerce. Thus the challenge of electronic commerce is to capture the information and knowledge in people's heads and place it in computer systems. Like people, the computer systems must be knowledge-based, adaptive and continue to learn from experience. Sound like Buck Rogers? Not really. Adaptive, reactive, and reflective task knowledge is the essence of a field of computer science known as distributed artificial intelligence (DAI), popularly known as intelligent agent technology. It is one of the most rapidly advancing computing technologies. Once codified into digital software agents, the task knowledge once isolated to a human specialist becomes available to an entire workforce, assisting the customer service representative at the call center, the salesman in the field and the customer himself with customer self-service (outsourcing customer service directly to the customer).
The company that masters eCommerce is the company that will master its industry in the emerging digital economy. Such mastery is not an end game, it is a continuing and unfolding process. Agility is the byword of success - an agile business empowered by agile information systems.
Evaluating and partnering with the emerging eCommerce software vendors should be a top business priority. Distributed architecture, open standards, and the appropriate use of agent technology are the key criteria for selecting eCommerce vendors, platforms and frameworks. To understand what we want computers and our software vendors to do for us in the realm of eCommerce, we can look at the real world of commerce. A successful eCommerce business strategy has three overall dimensions: 1. Develop eCommerce Strategies That Build on Existing Market Spaces, 2. Put Your House in Order With Enterprise Systems Architecture, and 3. Manage the Complexity with Intelligent Agent Technology.
1. Develop eCommerce Strategies That Build on Existing Market Spaces
To explore the functionality needed to participate in eCommerce, Figure 1 shows the three pillars of eCommerce that stand on the foundation of an existing market space: electronic information, electronic relationships, and electronic transactions.
Figure 1. Building Electronic Commerce
The electronic information pillar is, at least initially, the easy part. The Web is a global repository of documents and other forms of multimedia. Everyone can be a publisher of information with as little as a HTML enabled word processor, a free file transfer program, and a $19 a month account with an Internet service provider. Product and service information is easily accessed and displayed anywhere, anytime, giving the smallest of businesses a "Web presence." Forrester Research reported that by the end of 1996 about 80% of Fortune 500 had established Web presence compared to 34% in 1995 (Economist, May 10, 1997).
From this entry point, however, things can become complicated quickly. For example, an autoparts store that puts its catalog on the Web, will overwhelm the customer who must navigate the 40,000 or so awkwardly coded part numbers carried by even the smallest of autoparts stores. And that's not all. The customer needs to know if the part is in stock. If not, is the part available from the store's jobbers or wholesale distributors? Is the part available for restocking and, if so, when? Parts are can be substituted, and quite often they are superseded by newer parts.
Existing information must be translated into appropriate formats for the Web: HTML, XML, GIFS, JPEGS, RealAudio and so on. Tools for accomplishing such content publishing can be as simple as FrontPage, Powerpoint, or Word. Managing the content is the challenge, often requiring specialized document management software such as Domino, LockOut, Documentum and others. Catalogs are major endeavors, especially when the content one company wants to display comes from suppliers using different formats and data as well as in-house content. And, who wants to see an entire catalog? Most customers will only want the parts relevant to them and in the order they wish to view the information. Catalog management can become quite complex, especially when catalogs are virtual, combining information from several sources into a consolidated catalog that can be extended, inspected, resequenced, and subsetted. Making the right information available to the appropriate entities is another issue. Wholesale prices, for example, need to be available to retailers, but not to consumers. Thus catalogs need to know who is looking at them and reveal only authorized information.
Sherif Danish and Patrick Gannon's new book from CommerceNet Press, Building Database-Driven Web Catalogs, provides an excellent discussion of the sate-of-the-art, and describes the need for Marketing Product Data Management Systems to synchronize engineering, marketing, operations, and ERP systems so that CD-ROM, Web, print and custom catalogs maintain integrity. The book also introduces third party catalog aggregators including TPN Register, LLC (www.tpnregister.com), Acquion, Inc. (www.acquion.com), and WIZnet (www.wiznet.com). The need for real-time updates to aggregated catalogs along with the need for configurator systems that can build a product on the fly indicate that "behavior," not just data is essential to web catalogs of the future.
Search engines are of course important, but catalog information, to be effective, should be "pulled" by the customer to match individual preferences. This personalization capability requires profiles, a sensitive subject. Profiling is a hotly debated privacy issue, and to address the issue standards such as the Open Profiling Standard (OPS) are beginning to emerge. Profiling can be intrusive, using those little "Cookies" on your hard drive. Nonintrusive profiling can be accomplished in ways to insure privacy. Click streams can be captured, maintained, and analyzed on the eCommerce host system or "trusted" third party intermediaries can use their software proxies to ensure anonymity on the Web - a service sure to be vital to the widespread acceptance of eCommerce.
As this discussion reveals, there is much more to the electronic information pillar of eCommerce than simply publishing content. Dynamic publishing of information can require many tasks and intelligent decisions to be made in real-time. As we will see later, the use of intelligent software agents will be essential to handling these many tasks of creating, maintaining, and delivering content on the Web. And software agents will be needed to allow the user to intelligently navigate and interact with the information glut. While the graphical user interface (GUI) was a boon to personal computing by allowing direct manipulation of objects on the screen, task delegation (to software agents) is the next big step needed in user interfaces.
To be successful in eCommerce, a company must build the central pillar, Electronic Relationships. Building a site and hanging out a shingle on the Web, doesn't mean they will come. First and foremost the site must create the "buzz," much like Amazon.com has done in the book selling business. The site must be innovative, add value, provide information and interaction not otherwise available and create forums for opinion-building activities. In short, the site must build community and become the "port of entry" for commerce. Fruit of the Loom (FOL) understood this concept when it built its ActiveWear eCommerce site. It wanted to become the port of entry in its industry where everyone would come to transact business, tell their stories and display their wares. FOL positioned itself as the preeminent cybermediary by using its eCommerce systems to not only host its own site, but to provide their expertise, software and hosting to their distribution chain. Even the smallest mom-n-pop FOL distributor now can have a sophisticated site with all three pillars of eCommerce, thanks to FOL, without having to acquire hardware, ISP services, software, et al. And the distributors are not restricted to handling FOL products only. Customers can buy Hanes underwear from these distributors using the complete facilities of FOL. FOL exemplifies the business nature of eCommerce issues. Industry positioning was a well thought business strategy developed by the CEO and executive management, not an EDI programmer.
A Web site must not only get them to come, it must get them to come back, and come back regularly because it is so important to them. This is the essence of eCommerce. Unlike the mass media of print and broadcasting, the Web is an interactive medium. Interactive tools for building electronic relationships can be a simple as email. More advanced tools include list servers, news groups, bulletin boards, chat rooms, and in the future virtual spaces may be common such as those developed by Live Board and ActiveWorlds. Connecting people scattered around the globe who have common interests takes us once again to software agents where products such as Firefly use collaborative filtering technology to make the connections.
Two choices are available for determining customer interest: having the customer fill out an interest profile information or having the system analyze a user's activity trails through a web site. Such profiling processes can allow personalization and information filtering. Collaborative information filtering can be used to connect people with similar interests, can enable opinion making and, using advanced techniques such as artificial neural networks, uncover tacit information. Analysis of consumer buying behavior will become increasingly important as 1-to-1 marketing gains momentum on the web and as autonomous software agents offer the potential to intelligently mine for buying patterns from vast pools of customer information. Intelligent agents are a natural for customer data mining and can help to avoid the human SQL "query from hell" as well as uncover patterns humans would likely miss.
While data mining provides an opportunity for reflective, usually off-line, analysis of buying behavior, real-time profiling is essential as well. Forgetting about the Web for a moment, companies such as Catalina Marketing allow the supermarket clerk to print coupon offers directly targeted to the customer at the check-out lane, in real-time, at the point of sale based on the current shopping basket. Such pinpoint profiling accuracy in real-time is the key to marketing, adding value to individual customers, and cross-selling. Building customer relationships depends on both deep analysis of historical customer information and up-to-the-second real-time data - the winning combination for eCommerce.
Customer relationship management (CRM) is not just computer based. In fact the most important medium for CRM is today's call center, and companies that operate call centers should look to the knowledge and experience they have gained with this function to serve as the foundation for building electronic relationships. The integration of traditional call centers and Web-based customer service is inevitable. The two technologies will be integrated via Computer Telephony Integration (CTI) using IP telephony. Existing call center providers such as IBM's recently acquired Early, Cloud & Company are integrating the Web and CTI as additional customer "touch points." Such moves are essential to IBM's e-business strategy and provide an evolutionary path for users of such hybrid systems. Companies embracing the Web for CRM should include their call center team as cornerstone players in their initiatives.
The capability to handle Electronic Transactions is the third and least mature pillar of eCommerce. Perceived and real obstacles have to be overcome before electronic transactions become a widespread means of consummating binding contracts and making payments. Back to the Forrester report mentioned above, only 5% of the Fortune 500 companies with Web presence handle transactions on the Web in 1996. The major barriers include security on the Internet, lack of accepted standards for authentication and payments, non-repudiation, and general fear and uncertainty by consumers. Several commercial standards and offerings have already appeared including SET for credit cards, Open Buying on the Internet (OBI), Cybercash, ecash, digital wallets, microcash, InternetBill and ECXpert.
Business-to-business incarnations of eCommerce have grown more rapidly than business-to-consumer forms probably due to the widespread use of traditional Electronic Data Interchange (EDI). Internet forms of EDI are simply extensions of accepted technology, and the Internet Engineering Task Force (IETF) has already undertaken work this critical area, the Electronic Data Interchange-Internet Integration (EDIINT). The newly formed XML/EDI Group is aggressively promoting a standard that would open up business-to-business data exchange or "Web-based EDI" to millions of companies and individuals who would have never considered expensive EDI systems.
Currently, most eCommerce transactions such as retailer to consumer are simple - browse a catalog and make selection, then make a payment using a credit card. This kind of transaction in no way is reflective of the complex and nested transactions of real commerce. Transactions in the real business world are often long lived propositions involving negotiations, commitments, contracts, floating exchange rates, shipping and logistics, tracking, varied payment instruments, exception handling and termination or satisfaction. Commercial transactions have two phases, construction (information collection involving catalogues and brokerage systems to locate sources; agreement leading to terms and conditions through negotiation mechanisms; and engagement resulting in the "signed contract") and execution (configuration involving deployment across the group of participants in the transaction; service execution in context of the higher level contract and management of exceptions; and termination involving validation and closing the contract across all participants). Termination may be very long lived as contracts may include ongoing service agreements with online customer service delivery and other aspects of overall customer relationship management. In the world of eCommerce, traditional transactions are replaced with long lived, multi-level collaborations.
In a joint white paper of the Object Management Group and CommerceNet, Gabriel Gross, President of Centre Internet Europeen, summarized the current state of electronic commerce applications as "mainly limited to two functionalities: cataloging on one side and payment facilities on the other side. The [current] electronic commerce world is in practice a lot less sophisticated than real world commerce where several levels of interaction can take place between a potential client and vendor, and several levels of intermediaries can act or interfere."
What happens when eCommerce matures to reflect the real and complex world of multi-seller, multi-buyer commerce? The real world of eCommerce will require building digital simulations of real world open markets. As shown in Figure 1, Open Market Processes provide the capstone for an enterprise wanting build information systems capable of participating in the coming digital economy. Such an open market is illustrated in Figure 2. Gross explains, "… an open electronic market infrastructure will act as the foundation of an ecosystem where actors will be able to join in and play the roles they want. The roles of the individual actor will depend on the situations, on the other actors he is faced with, or on his motivations of the moment."
Figure 2. Open Digital Markets
A given actor may play multiple, simultaneous, and dynamic roles: customer (the one who pays), consumer (the one who receives), merchant (the one who gets paid), and provider (the one who delivers). And there is more to these basic roles when considering the roles of broker, aggregator, referral service and other forms of intermediation along with the supporting roles of the banker, credit card provider, shipper, insurer and other third parties.
Open market processes require interoperation of computer applications and consistent protocols and formats for information interchange. The complexity of building such virtual commerce places mandates a distributed computing paradigm based on standards if there is to be any hope of interoperability. The ultimate purpose of standards is the development of consistent business semantics that can be used by all participants - a common language of digital commerce. These semantics provide commonality to names and relationships of processes, workflows and data across value and supply chains. A new standard for defining and naming data on a Web page was adopted by the World Wide Web Consortium (W3C) in 1997. The eXtensible Markup Language (XML) will likely revolutionize the Web by allowing structured data - with standard names and consistent semantics - to be moved around the Web in a simple, straight-forward manner.
XML is, however, no "silver bullet" and in some ways is little more than the reintroduction of the "unit record concept" introduced with the punched card in the 1950's where chuncks of data (fields) were tagged with names giving us attribute/value pairs bound together as a standalone document (record)! After all, XML is simply text (ASCII) data and must have links to a powerful, underlying object infrastructure to handle the adaptive business processes and workflows needed for eCommerce. The truly hard part is to gain global agreement on the semantics - an effort that has eluded information systems designers since the introduction of centralized corporate data bases in the 1960s. Ask any experienced Data Administrator and they will tell you they cannot even get departments within the same company to aggree on data names, much less their meaning. Thus XML is simply an enabler, not a guarantor, of the consistent business semantics needed for eCommerce. Some industries will do better than others in establishing semantic consensus. Groups such as the XML/EDI Group may have early success as they are dealing with an established body of inter-enterprise business semantics. XML is discussed further in the next section.
Without consistent business semantics, processes and workflows cannot be shared across multiple, distributed, and disparate businesses. Sound complicated? It is. The task knowledge needed for such inter-enterprise activity (adaptive business processes and workflows) overwhelms today's software paradigms. Once again, the use of intelligent agent technology which is based on task level knowledge and knowledge sharing standards such as a simplified version of the Knowledge Interchange Format (KIF) will be essential to the future of open digital markets. Collaborative multi-agent systems are central to building open, digital commerce places as illustrated in Figure 2. The figure shows several enterprises convening around a commerce transaction and shows the multiple simultaneous roles of Enterprise A (for example, as a merchant and provider).
Trading process models being developed for eCommerce are fundamentally the same as traditional commerce processes. Core processes include pre-sales and sourcing, order placement, supply chain management, settlement and post-sales support. Coordination of these processes is essential and pioneers in developing open markets turn to coordination theory such as that being researched at MIT's Sloan School under the direction of Dr. Thomas Malone. Coordination is also at the heart of designing and building multi-agent systems. The collaborative working arrangements and processes of today's real commerce will eventually be simulated in the knowledge-based software of open eCommerce.
2. Put Your House in Order With Enterprise Systems Architecture
Enterprise computing is about consolidating and harmonizing the many islands of disparate information and systems scattered throughout an enterprise into a unified whole. The goal is to streamline business processes and enable outward-facing information systems. The attention given to enterprise computing over recent years is a result of the business process reengineering revolution which, in turn, was enabled by information technologies such as client/server computing. Through some hard learned lessons, corporations now know that it is not enough to wire together machines through a network using a client/server architecture. What is missing is a coherent information model as well as a technology architecture. This is where object-orientation comes into the picture.
Objects are chunks of software that reflect real things in the real world: customers, employees, orders, shipments, and so on. Objects combine their processes and their data into a single entity in such a way that the integrity of the object is assured by the object itself. This is in contrast to the relational database model typical of client/server architectures where data is isolated from the processes that manipulate it. Such processes can be scattered across an organization and many integrity and complexity problems result. Companies that tried to criss-cross connections among the many relational databases and processes that use them ended up with incomprehensible spaghetti architectures and were not successful in creating a unified enterprise information infrastructure.
Distributed object computing is now recognized as the way forward in building enterprise information architectures. Objects communicate to one another, to users, and other systems by presenting interfaces showing the services they can perform. To ask an object to do something, it is sent a message requesting a service. In essence, using objects to build information systems is like building a simulation that includes objects representing the people, places, things and events that are found in the business setting or domain. Four key advantages result. Objects reflect the real world and thus greatly enhance understanding and communication among systems developers and business people. Objects are stable as part of the object's internals or its interfaces can be changed without affecting other parts. Objects can help achieve software reuse as they can be extended through mechanisms such as inheritance without rewriting the entire object. And, essential to enterprise computing, objects reduce complexity as programmers do not have to understand the internals of the object. They do not need to know how an object works internally, only what the object is and to what messages it responds.
Distributed object computing has evolved rapidly over the past five years or so. Early uses of this computing paradigm dealt with system or "technology" objects. A printer is a simple technology object. A programmer does not need to "program" a printer any more. Instead the programmer sends the printer object a message requesting that the object take care of printing the current document. Traditional procedural programming required that programmers know all about programming printers, carefully writing each instruction to handle line feeds, tabbing and so on.
While technology objects are necessary they do not address the business applications or business semantics per se. What is needed is a higher level of abstraction that allows information systems developers to work with objects that represent real business entities and processes. At this level of abstraction powerful business information models can be designed using object-oriented concepts. Instead of handling systems tasks such as printing, business objects can handle the tasks of business processes and activities while suppressing the details of the underlying technology objects. The technology details are needed, of course, but the internals of the business object take care of theses matters by sending appropriate messages to the underlying technology objects. As the theory and practice of business objects continues to evolve, business people will instruct computers in business languages, not computer programming languages.
In addition to allowing corporations to work with business information models and semantics, the distributed object paradigm is essential to tying together disparate islands of information throughout an enterprise and across enterprises - the virtual corporation. The secret is that the objects need not be on a single machine. Objects can exchange messages across networks as easily as within a local computer system. This interoperation is possible among objects programmed in different languages (COBOL, C++, Java); existing heritage or legacy systems can be "wrapped" to appear and function as objects; and the platform can be of no matter, a mainframe, a Unix server, or a PC. An object request broker or ORB makes all objects on the network appear just as though they reside on the local machine. The Object Management Group, an 800 plus member organization, has developed the standards needed for truly open systems, the Common Object Request Broker Architecture (CORBA). Defacto standards also exist, but Microsoft's DCOM requires all participants to be selected versions of Microsoft only operating systems, and Sun's Remote Method Invocation (RMI) initially spoke Java language only and couldn't talk to C++ or Smalltalk objects until it was restructured above OMG's Internet InterOrb Protocol (IIOP). Fortunately bridges are being build to cross the chasms of these distributed computing standards.
Distributed object computing sets the foundation for both enterprise computing and the inter-enterprise computing that is essential to open processes of eCommerce. In fact, it is the distributed object paradigm that has made it possible for eCommerce to aspire to open digital markets. Like business objects in general, the interoperation of open market processes must rise above technology interoperation ("above the ORB") to a level of abstraction where business information models can interoperate. In other words, common business semantics must be available to all participants in an open digital market.
Like the little inscription on your car's rear view mirror says, "objects are closer than they appear." OMG's Common Object Request Broker Architecture (CORBA) has made system and technology interoperation available for a number of years now, and many mission critical business systems and applications have already been developed by Fortune 1000 companies with staffs who have climbed the learning curve of distributed object computing, handling low level technology objects along with business functionality. The interface definition language (IDL) specification allows programmers to write and publish interfaces that can be used by objects anywhere at anytime. To date, however, developers must master a mix of business object semantics and the underlying technology objects which require low level plumbing in order to build complete business solutions. What is needed is a business object component model that supresses the complexity of the underlying systems technology, thus providing a clear separation of concerns. With such a model, business solution developers can assemble and tailor prebuilt business components into complete solutions.
The exciting stuff in the object world is just now in the process of being standardized by the OMG, the Business Object Component Architecture (BOCA), with its Business Object Facility (BOF) and Common Business Objects (CBO). Business object components suppress much of the low level programming needed in the original CORBA approach and allow applications to be assembled or wired together (by non-experts) from the business components. How all this fits together is shown in Figure 3.
Figure 3. Business Object Component Architecture
As shown, CORBA is the foundation for distributed objects and provides objects not only transparent access to other objects, but also infrastructure objects known as CORBAservices and CORBAfacilities. Such common facilities and services allow programmers to "plug into" standard preprogrammed services such as transaction processing, permanent storage of objects, event handling services and the like. They may be thought of as the essential computing services, and operate similar to the way the printer object works - programmers ask for the services without having to know the internal workings of the objects that deliver them. These low level systems services call on the actual technology infrastructure that may consist of wrapped legacy systems as well as services provided by native objects.
Stepping one giant step above the CORBA infrastructure is OMG's new Business Object Facility (BOF) that may be considered as a run-time ORB for business object components. The BOF supports the semantics for business rather than technical interoperability. The BOF is an infrastructure that supports "plug and play" business application components. These components can be assembled or "glued together" to rapidly develop distributed applications. The Component Definition Language (CDL) allows component builders to define the interface to business objects with business semantics, rather than lower level system interface definitions specified in CORBA's Interface Definition Language (IDL). In other words, CDL is a higher level of business abstraction built on top of IDL.
The World Wide Web is extremely complex under the covers, but it is simple for a business user to browse the Web. The telephone is easy to use while telephony and telecommunications are very complex. From a software development perspective, complex is simple, simple is complex. As shown in Figure 3, technologists are component builders and deal with the complexity of the underlying information systems and technology infrastructure. Computer specialists are the toolsmiths for building reusable components. Because business objects provide the abstractions needed for building high level components that interoperate, business solution developers are able to assemble applications using business constructs and semantics that insulate them from the underlying complexity. The ultimate language of applications development will be the "language of business," not "the language of computers," and eventually business people will develop their own information systems solutions to business problems.
The component paradigm will forever change "programming." Application programmers who make up the bulk of today's commercial IS shops will simply go away. Where will they go? They will work for software vendors who create components. Microsoft, SAP, Baan, SSA, IBM, Oracle and most other software vendors are working hard to bring about the shift from writing code to component assembly. The components will not be delivered to corporations as a big pile of parts and pieces. Instead the components will be preassembled into industry specific application frameworks. These frameworks represent applications that are almost, say 60% - 70%, complete. The task of solutions developers will be to customize the components of the frameworks to meet the unique needs of the specific company.
Solution developers will concentrate on the unique character and knowledge of the company that provides competitive advantage. Much of the extension and tailoring of components will focus on the user interfaces and will involve both graphical and task-centered customization. In the long run, corporations will no longer design and code their applications. Instead they will buy component-based enterprise application packages. These packages, however, will not be the complex, hard to understand and configure packages we see in today's ERP world. They will be frameworks based on distributed object architectures that allow individual components to be mixed and matched even though they are built by a variety of software vendors. Hello frameworks, good bye compilers. Hello solution developers, good bye corporate programmers. As a precursor to this trend, notice the new phenomena of teaching vendor specific courses in business schools. SAP and other ERP vendors are aggressively implanting their software based on case studies into business (not computer science) curriculums at both the graduate and undergraduate levels. The result will be business graduates who can build information systems from frameworks. The programmer as "translator" between the business and the computer will fade into history.
What language will solution developers speak? CommerceNet is working on the Common Business Language (CBL) as one such language for gluing together eCommerce components in their evolving eCo System architecture. Other high level tools such as VisualAge for Java serve as component assembly means and suppress the details and complexity of the underlying technologies and systems. Sure there will be new languages to master, but the task will be mastering the semantics of business, not the computer.
At the higher level of Figure 3, notice the Common Business Objects (CBOs) that represent objects of interest across industry bounds - customer, shipping document, account, order, and so on. Using these CBOs, industry specific business objects and framework specifications are being developed by various consortia and standards groups such as the OMG's domain task forces: Manufacturing Domain TF, Telecommunications Domain TF, Financial Domain TF, CORBAmed TF, and the Transportation Domain TF. Meanwhile, the Open Applications Group (OAG) is forging standards for applications interoperability among ERP vendors. ACORD is taking on the standards task in the insurance industry, the Supply Chain Council is addressing supply management, RosettaNet is focused on the IT industry supply chain, and POSC is addressing the standards needed in the oil and gas industries. Through such groups consistent business semantics are defined and evolved that enable the building of open markets and processes.
OMG's Electronic Commerce TF is tackling the domain of electronic commerce business objects and service architecture for open electronic markets, the Electronic Commerce Reference Model shown in Figure 4. The model is of sufficient depth to address such issues as "policy management" which affects agent behavior and enterprise system rule enforcement, and reaches deeply into the interactive relationship between participants such as customer and supplier (and their respective agents). Negotiation may equally occur over agreement on policy in order to establish a common, temporary domain. This concept of "dynamic policy domains" built using sophisticated policy management tools opens up new markets in third-party policies, used for the duration of a transaction and then discarded. For information and status of this vital work, visit http://www.osm.net/about/library.cfm.
Figure 4. OMG's Electronic Commerce Reference Model Architecture
While somewhat differing in approach to interoperation among commerce participants, CommeceNet is developing its eCo System architecture around a document-centric paradigm using the new Web standard, the eXtensible Markup Language (XML). XML describes structured data packages that move around the Web as easily as HTML does today. XML is a native Web approach that enables extensible data-exchange formats, and gives industries the flexibility to create their own data tags to develop a shared Internet file system. With links to processes and objects containing behavior, XML can make commerce applications such as Electronic Data Interchange (EDI) available to small to mid-sized companies that have not been able to afford such systems. The eCo System architecture does not preclude OMG's or other object models due to linking facilities that allow behavior to be added to XML documents. Quite the opposite, eCo System needs access to such deep and adaptive object infrastructures to handle the business processes and workflows inherent in real commerce. In fact, the World Wide Web Consortium (W3C) is developing the next generation of the Web to incorporate an underlying distributed object model - HTTP-NG. For an informative report, "Towards a Web Object Model," visit http://www.objs.com/OSA/wom.htm. Companies seeking architectural direction will keep a close eye on these standards activities, and those intent on leading their industries will participate directly in developing the standards.
Putting their information systems house in order (so they can integrate their internal islands of information into a unified whole and interoperate with their suppliers and customers) should be the highest priority of CEOs and other business leaders. This is the stuff of business strategy for the future. Not only handling, but embracing change is the foremost criteria. Distributed object computing and enterprise system architecture provide the blueprints.
Building technology alliances will be key. Leading software vendors are hard at work to establish leadership in this brave new world and deliver solutions today, not in some distant future. For a quick survey visit the following web sites: IBM's San Francisco Project (http://www.ibm.com/Java/Sanfrancisco/prd_summary.html), Broadvision's One-to-One (http://www.broadvision.com), OSM Sarl's Business Agents (http://www.osm.net), and EC Cubed's ecWorks™ (http://www.eccubed.com), and webMethod's B2B (http://www.webmethods.com) . Visit commerce.net and omg.org to find other leading firms building the future of eCommerce.
The goal of enterprise systems architecture is to design information systems that have the intelligent interoperability and agility demanded of them in the unfolding world of eCommerce. The tasks are to build a solid architectural foundation (that preserves existing heritage systems) and directly participate in the work of setting the standards.
3. Manage the Complexity with Intelligent Agent Technology.
From our discussion so far we can conclude that the world of eCommerce is not only complex, it is laden with a multitude of tasks. And most tasks are not just straight forward "do this, do that" lists. Successful computer systems underlying eCommerce require judgment and the knowledge of experts such as buyers, contract negotiators and marketing specialists. Agent technology authority Jeffrey M. Bradshaw explains, "Recent trends have made it clear that software complexity will continue to increase dramatically in the coming decades. The dynamic and distributed nature of both data and applications require that software not merely respond to requests for information but intelligently anticipate, adapt, and actively seek ways to support users. Not only must these systems assist in coordinating tasks among humans, they must also help manage cooperation among distributed programs."
An open digital market is a complex and non-deterministic system, often producing results that are ambiguous and incomplete. Auctions and contract negotiations are obvious examples. Knowing when to sell in an auction requires judgment based on estimates of demand and price elasticity - the kind of judgment traders use in traditional financial and commodity markets. In addition, configuring and reconfiguring software to participate in different, often simultaneous roles in digital markets could not be accomplished using the old fashioned way of manually configuring software. Further when a price change occurs in a product or product component, real and virtual catalogs in each trading place they can be found must be updated accordingly. Add to this the need to profile customer preferences and create communities of like interest, humans would be simply overwhelmed by the zillion knowledge-based tasks inherent in eCommerce. What is needed is software that can take on these tasks. Such software needs to be "smart" enough for the tasks at hand, simple stimulus-response for handling simple tasks or highly deliberative for mining databases of customer information. Such software must be able to "play the game" of the markets in which they compete. It must be goal seeking just a salesman is goal seeking in the game of competitive selling. The essence of such software is taking task knowledge now contained in the heads of the traditional commerce workforce and codifying it into eCommerce software systems. This is not some Orwellian vision of eCommerce, but characteristics even a casual observer could make of the requirements open digital markets.
Electronic games, automated traders in the stock market, Deep Blue playing world class chess, case based reasoning determining credit approvals at America Express, and expert systems of all sorts are well known and used today. They deploy various techniques and knowledge from the field of artificial intelligence (AI), a field that was overhyped a decade ago and fallen from grace in the commercial computing press. Quietly, however, AI has continued to play a significant role in many leading information systems. Its use has been limited due to its complexity, monolithic designs and lack of knowledgeable systems developers. On the other hand and like other technologies such as object technology, AI is making a very strong comeback as it is now crucial in non-deterministic systems such as workflow, data mining, production scheduling, supply chain logistics, and most recently, eCommerce. Its new form is not the monolithic AI systems of the past, but distributed artificial intelligence (DAI), popularly known as intelligent agent technology.
Going one step beyond distributed object technology, intelligent agents are poised to transform the way we model the enterprise and build information systems. Both object and agent paradigms address change and complexity. Intelligent agent technology is the next logical step in moving the object paradigm forward and overcoming some of its shortcomings in eCommerce.
Thanks to the dramatic growth of the Web, computing literature is replete with discussions of software agents. Like objects, however, the term is weary from overuse and misuse. Various names have been used - knowbots, softbots, personal assistants, and smart computer programs. Cutting through the hype the importance of agent technology, varying in "intelligence" from simple stimulus-response sensors to highly deliberative agents capable of learning, is the central feature of software agents, task capabilities. Robust eCommerce systems being built today simply could not function without being able to delegate to software the multitude of tasks that would otherwise be left to armies of people to handle.
Backing away from technology for a moment, the everyday term, agent, provides a starting definition: "one who acts for, or in the place of, another." A software agent is a software package that carries out tasks for others, autonomously without direct intervention by its master once the tasks have been delegated. The "others" may be human users, business processes, workflows or applications. A basic software agent has three essential properties: autonomy, reactivity, and communication ability. The notion of autonomy means that an agent exercises exclusive control over its own actions and state without control of some external software entity. Reactivity means sensing or perceiving change in their environment and responding through effectors. And, even the most basic software agents have the ability to communicate with other entities (human users, other software agents, or objects) giving them various forms of social ability.
Mobility is often mentioned in discussions of essential agent qualities. The notion of mobile agents roaming the web doing the bidding of their masters is all over the computing literature. Debates rage on Internet discussion groups over the nature and necessity of mobility in agents. The need for roaming mobility is highly exaggerated and will more than any misconception create fear and uncertainty that stunts the adoption of agent technology in eCommerce. As with objects, messaging, not roaming is the key. Agents are anchored at nodes of an information network, and although they can be mobile, more than likely they will be stationary or shipped to cooperating information nodes only when it makes more sense to place the functionality at the node versus using network messages: load balancing and conserving network bandwidth. "Parachuting" in where they are needed is perhaps a better metaphor than agents roaming, pilfering and plundering information systems on a network. Mobile agents and mobile objects are important in distributed systems, but care should be taken when using these notions that can trigger fear in the uninitiated.
"Agenthood" can be measured along two axes: agency and intelligence. Agency is the degree of authority and autonomy given to the agent as it interacts with its user and other agents in an environment. Intelligence is the degree of reasoning and independent learning abilities of the agent. Intelligent agents support a natural merging of object orientation and knowledge-based technologies that have grown out of the field of artificial intelligence. What are these artificial intelligence contributions and why are they important to eCommerce? Even "dumb" software agents incorporate simple Boolean logic: "If the stock price drops to $50, notify the user." If only the world of eCommerce was so absolute, our discussion would be complete. But what we need are knowledge-based agents that can reason under uncertainty, goal-driven agents that can plan, and agents that learn. To pursue these characteristics we turn to the artificial intelligence technologies of fuzzy systems, artificial neural networks, and evolutionary processes such as genetic algorithms.
If an agent needs to reason with imprecise or incomplete information or linguistic variables, it will need to incorporate fuzzy logic. Fuzzy systems extend Boolean logic to handle degrees of truth that fall in between being completely true or completely false, for example the variable 'Age' whose linguistic values are young, old or middle-aged. Such unsharp boundaries are commonplace in the domain of commerce - after all, cheaper, better and faster are not absolute values. Fuzzy systems enable the building of intelligent agents that can be used for business process monitoring, scheduling, planning, forecasting and natural language processing.
Artificial neural networks are composed of groups or layers of interconnected processing elements analogous to the neurons of the brain. Each neuron has one or more inputs and one output, and uses the weighted sum of the inputs to produce a value for output or not - determining if the neuron "fires." What is important is not a single neuron, but groups of them whose initial weights are set to initial values. Inputs are applied, and comparing the output to some goal results in measurable errors. Then the weights are reset and the process repeated until the goal is reached. The resetting of the weights is done according to a set of rules, and the backward propagation of errors through the net provides feedback that can "train" the net. In this way the net organizes itself, it learns. Using noisy or incomplete historical or real-time market data a neural net can train itself to learn patterns that lead to critical market conditions, predict trends, and enable agents to improve their performance of tasks over time.
Evolutionary computing techniques can give agents the ability to evolve their reasoning and behavior in environments where conditions are dynamic, such as multi-buyer, multi-seller market places. Techniques such as genetic algorithms "evolve" software processes through simulations constrained by the laws of reproduction, natural selection, and mutation. Individual processors that prove unfit are eliminated or "die out" through a process of natural selection. Parent solutions to a problem produce offspring that change or "mutate" the parent's behavior. Then the offspring are subject to their fitness as a solution. Evolutionary computing techniques are suited to optimization problems such as resource and job scheduling and production control needed in most supply chain environments. These techniques are relevant in situations where there are many interacting variables that can result in many possible solutions to a given problem. The intelligent agent using such techniques can produce optimal solutions from among all possible solutions. Physical distribution of finished goods, task scheduling in workflows, cash flow optimization, concurrent design and engineering, and dynamic network configuration are typical applications.
To sum up these "soft" computing approaches in eCommerce we turn to the notion of taking a vacation. Today several resources are available on the web to reserve and book your flights, local transportation and lodging (Travelocity, Expedia, theTrip). Price and availability are the two criteria used to match a user's needs. But what if we wanted to design an intelligent travel agent that could optimize our vacation. For example the lowest airline fare may have us taking an all night red-eye with multiple connections instead of what could have been a 1 hour nonstop flight for a few dollars more. Using price alone and arriving exhausted may not be the optimal start to a the vacation we had in mind. Our intelligent travel agent will have to look at many variables including subjective fuzzy variables to become a really good travel agent.
In their book, Developing Intelligent Agents, Knapik and Johnson paint the scenario, "Where a nonintelligent agent may compare airfares to find the lowest fare to a specific destination, an agent that provides successive generations of better and better vacation solutions would be much more adept at pleasing the user. You could use a genetic algorithm tool to encapsulate the process, which lets a vacation-planning agent not only formulate a vacation plan, but find the best one for its user's interest, available funds, time, and so on. This would involve dynamically modeling many possible vacations and ranking them based on current criteria. The agent may find that there is snow at a user's favorite ski resort, but that lift-ticket prices just went up. It may also detect additional funds in the user's vacation account, so a little fancier vacation might be possible. As a wide range of vacations are thus modeled and judged for maximum expected utility, a short list of winners eventually emerges, and can be evaluated by the user. If none on the short list is satisfactory, the user can kick the whole process off again, whereby the parameters would be combined or mutated by the genetic algorithm to arrive at a new set of vacation solutions." This illustration of soft computing with appropriate and multiple intelligences weaved throughout software is the future of eCommerce. All possible intelligence, however, is not simply thrown into all possible agents, and there is no single, all-knowing intelligent agent. It is the building in of the right kind of intelligence for the right purpose that is the key to successful design.
"Intelligent agents" build on the three pillars of a basic software agent. They add the ability to plan and set goals, to reason about the effects of actions, and to improve their knowledge and performance through learning. The architecture of contemporary intelligent agents is object-oriented and is essentially an extension of an object. Objects encapsulate data (attributes) and procedures (methods). Add to this the intelligences of fuzzy systems, artificial neural nets and evolutionary computing techniques, and an object becomes an intelligent agent. Intelligent agents can act autonomously, in the background of mainstream applications, to perceive their environment through sensors and act on the environment to perform and optimize their tasks: information retrieval, monitoring and notifying, information filtering, data mining, coaching, negotiating, configuring and so on - all the multitude of tasks needed in a robust eCommerce environment. The following table highlights the diverse requirements and uses of intelligent agent technology in eCommerce.
|
The New Book for the New Millennium 
A CEO's Guide to eCommerce Using Intergalactic Object-Oriented Intelligent Agents by Peter Fingar discusses the new electronic commerce and the intelligent agents that will make it possible and what CEO's need to be doing and thinking now to prepare themselves and their companies.
