collaborative commerce and knowledge management

&

Collaborative Commerce andKnowledge Management

Bhavani Thuraisingham1, Amar Gupta2, Elisa Bertino3 andElena Ferrari3

1MITRE2MIT3University of Milan, Italy

This paper describes collaborative commerce (c-commerce); it essentially combinese-commerce, knowledge management and collaboration to carry out transactions and otheractivities within and across organizations. We first discuss the building blocks forc-commerce. Then we describe models and federated architectures. Next we analyze thestrategic role of knowledge management for c-commerce as well as discussing managerial andbusiness implications. Finally, we provide directions for c-commerce. Copyright # 2002 JohnWiley & Sons, Ltd.

INTRODUCTION

Three closely related information technology dis-ciplines have emerged during the past decade. Theyare collaborative computing, knowledge manage-ment, and electronic commerce (e-commerce). Col-laborative computing enables people, groups ofindividuals, and organizations to work togetherwith one another in order to accomplish a task or acollection of tasks. These tasks could vary fromparticipating in conferences, solving a specificproblem, or working on the design of a system.With the advent of the web, it is now possible fororganizations to effectively manage their know-ledge and experiences. That is, knowledge man-agement is about using knowledge as a resourcefor an organization and subsequently creating aknowledge organization with knowledge workers.Electronic commerce (e-commerce) is one of thefastest growing application area for the WorldWide Web. Essentially, it is about carrying outtransactions on the web such as buying and sellingproducts, as well as carrying out various activitiessuch as training, procurement, and supply chainmanagement.

While the three disciplines have evolved some-what independently over the past decade, they

have a lot to contribute to each other. Effectivecollaboration is key to knowledge managementand good knowledge management practices areessential for a successful e-business organiza-tion. More recently, organizations are going bey-ond collaboration, knowledge management ande-commerce. These organizations are collaborativelycarrying out transactions on the web and goingbeyond organizational boundaries. That is, theorganizations are autonomous and yet cooperatewith each other. This form of collaboration hascome to be known as c-commerce (or collaborativecommerce) (Chen, 2000). Essentially, organizationshave to work across international boundaries,collaboratively carry out transactions, as well asshare and collaborate on activities and projects.Such collaborations can span across not onlyorganizations but also continents. The participantsmust be able to share information, but at the sametime protect their privacy as well as sensitiveinformation. Disparate heterogeneous informationsystems have to be integrated to form federationsand subsequently work effectively within as wellas across organizations for c-commerce.

This paper focuses on models and architecturesfor c-commerce, and then analyzes the role ofknowledge management. In the next section we

Knowledge and Process Management Volume 9 Number 1 pp 43–53 (2002)DOI: 10.1002 /kpm.132

Copyright # 2002 John Wiley & Sons, Ltd.

first discuss the building blocks for c-commerce.Then in the third section we discuss businessmodels for c-commerce and introduce an example,which we will use throughout the paper. In thefourth section we describe how c-commerceenables organizations to work across boundariesof various types. In particular, federated archi-tectures for c-commerce will be discussed. The roleof knowledge management will be stressed in thefifth section. We also briefly address business andmanagerial implications for knowledge manage-ment and c-commerce in this section. Finallywe discuss the directions and challenges forc-commerce.

BUILDING BLOCKS FOR C-COMMERCE

The building blocks of c-commerce are essentiallye-commerce, knowledge management and col-laboration. This section discusses these buildingblocks.

E-business and e-commerce

By definition e-commerce involves carrying outcommerce on the web, which includes buyingand selling of products. The term e-business isabout carrying out any business on the web andis broader than e-commerce. Various types ofcorporations are now in e-business, includingcorporations that provide consulting as well assolutions and products such as IBM, and smallercorporations such as the dot-com companies. Someof these smaller corporations can connect con-sumers with healthcare providers, lawyers, realestate agents and other professionals who provideservices of various kinds. Consulting companiesmay come in and assess the state of a corporation’sbusiness practices and advise it on how to develop

e-business solutions. One of the latest trends is toprovide fully integrated enterprise resource man-agement and business process reengineering cap-abilities on the web. A strong business component isessential for e-business. Technology will provideonly the tools to make e-business more efficient.Various aspects of e-business are illustrated inFigure 1 (see also the discussion in Thuraisingham,2000).

Two paradigms, which we can consider to bemodels for e-commerce, are emerging. They arebusiness-to-business e-commerce and business-to-consumer e-commerce. As its name implies, busi-ness-to-business e-commerce is all about twobusinesses conducting transactions on the web(see Figure 2) (Tracy, 2000). We give some exam-ples. Suppose Corporation A is an automobilemanufacturer and needs microprocessors to beinstalled in its automobiles. It will then purchasethe microprocessors from Corporation B who manu-factures the microprocessors. Another example is

Figure 1 E-business and its components

Figure 2 Business-to-business e-commerce

Knowledge and Process Management

44 B. Thuraisingham et al.

when an individual purchases some goods, suchas toys, from a toy manufacturer. This manufac-turer then contacts a packaging company viathe web to deliver the toys to the individual.The transaction between the manufacturer andthe packaging company is a business-to-businesstransaction. Business-to-business e-commerce mayalso involve one business purchasing a unit ofanother business or the two businesses mergingtogether. The main point is that such transactionshave to be carried out on the web.

Business-to-consumer e-commerce is when aconsumer such as a member of the mass popula-tion makes purchases on the web (see Figure 3). Inthe toy manufacturer example, the purchasebetween the individual and the toy manufactureris a business-to-consumer transaction. Business-to-consumer e-commerce has grown tremendouslyduring recent years. While computer hardwarepurchasesis still constitute the leading applicationfor e-commerce transactions, purchasing toys,apparel, software, and even groceries via the webhaves also increased. But many experts believethat the real future will be in business-to-businesstransactions, as this will involve much greateramounts of transactions. Furthermore, for organ-izations to work across boundaries, we needeffective business-to-business c-commerce. Accord-ingly, we focus on the latter segment in this paper.

Knowledge management

While e-business and e-commerce enable businesstransactions to be performed on the web, oneneeds effective knowledge management and col-laboration techniques for organizations to workacross organizational boundaries (Winch, 2000).

Knowledge management is the process of usingknowledge as a resource to manage an organiza-tion. It could mean sharing expertise, developing alearning organization, teaching the staff, learn-ing from experiences, as well as collaboration(Saunders, 2000). Knowledge management essen-tially changes the way an organization functions.Instead of competition, it promotes collaboration.This means managers have to motivate employees,to share ideas and collaborate, by giving awardsand other incentives. Team spirit is essential forknowledge management. People are often threat-ened imparting knowledge to others as they feeltheir jobs may be on the line. They are reluctant toshare their expertise. This type of behavior couldvary from culture to culture. It is critical thatmanagers eliminate this kind of behavior not byforcing the issue but by motivating and educat-ing the staff in all the benefits that can occurwith good knowledge management practices; thisaspect becomes especially important when oneis dealing with situations involving borders ofdifferent types.

In the collection of papers on knowledgemanagement by Morey et al. (2001), knowledgemanagement is divided into three areas as shownin Figure 4. These are strategies such as building aknowledge company and making the staff know-ledge workers; processes such as techniques forknowledge management including developing amethod to share documents and tools; and metricsthat measure the effectiveness of knowledgemanagement. Organizational behavior and teamdynamics play major roles in knowledge manage-ment.

While knowledge management practices haveexisted for many years, the web has really promoted

Figure 3 Business-to-consumer e-commerce


Collaborative Commerce and Knowledge Management 45

knowledge management. Many corporations nowhave an Intranet, which is a very powerful know-ledge management tool. Thousands of employeesare connected through the web in an organization.Large corporations have sites all over the worldand employees are becoming well connected withone another. Email can be regarded to be one ofthe early knowledge management tools. Nowthere are many tools such as search engines ande-commerce tools. With the proliferation of webdata management and e-commerce tools, know-ledge management will become an essential part ofthe web and e-commerce. Figure 5 illustrates theknowledge management activities on the web suchas creating web pages, building e-commerce sites,sending email, and collecting metrics on webusage.

Collaboration

While e-commerce enables organizations to carryout transactions, and knowledge managementcaptures the knowledge of an organization, westill need a third component to enable organiza-tions to work together and yet be autonomous.That third component is collaboration. As men-tioned above, collaborative computing enablespeople, groups of individuals, and organiza-tions to work together to accomplish a task or acollection of tasks. These tasks could vary from

participating in conferences, solving a specificproblem, or working on the design of a system.Specific contributions to collaborative computinginclude the development of team workstations(where groupware creates a shared workspace sup-porting dynamic collaboration in a work group),multimedia communication systems supportingdistributed workgroups, and collaborative com-puting systems supporting cooperation in thedesign of an entity (such as an electrical ormechanical system). Several technologies includ-ing multimedia, artificial intelligence, networkingand distributed processing, and database systemsas well as several diverse disciplines such asorganizational behavior and human computerinteraction have contributed significantly towardsthe growth of collaborative computing.

While early collaborative computing systemsfocussed on small teams working together, colla-boration has taken on a different dimension with theadvent of the web. Organizations can now collabo-rate across international boundaries, without mak-ing any a priori arrangements, and performmutual and joint transactions in a near-real-timebasis (Caldwell et al., 2000).

Putting the building blocks together

One of the key aspects here is how does e-commercediffer from c-commerce and what is the role of

Figure 4 Knowledge management components

Figure 5 Knowledge management on the web



knowledge management in c-commerce? Whilee-commerce may assume an environment wheresellers sell their products on the web while buyerspurchase these products, c-commerce assumes anenvironment where organizations have to collabo-rate with each other to carry out transactions. Webcollaboration is an essential part of c-commerce.For example, groups of designers and manu-factures have to collaborate with one another inorder to come up with a new design or product.C-commerce assumes that organizations are auto-nomous but collaborating entities that carry outbusiness with each other.

Knowledge management plays a key role inc-commerce. It captures the best practices andexperiences from previous transactions and makesuse of them effectively to carry out c-commerce.We will elaborate on the role of knowledge

management later. Figure 6 illustrates the integra-tion of the building blocks for c-commerce. Moredetails on the building blocks can be foundin Flynn (2000), Ginsburg and Kambil (1999),Malhotra (2000), Rehfeldt and Turowski (2000)and Scott (2001).

MODELS FOR C-COMMERCE

In order for organizations to carry out c-commercecollaboratively, one needs suitable business models.Seller to buyer relationships in c-commerce can takeone of many forms. Figure 7 attempts to illus-trate this. In a tightly coupled model, one partydominates the business practices and the othersconform to them. Because the dominating party isso large, it is either the dominant seller or the

Figure 6 Putting the building blocks together

Figure 7 Models for c-commerce



dominant buyer in the market. Therefore, it hasthe power to dictate what system will be used toconduct the transaction. Tightly coupled relation-ships can take the form of a one-to-one, a many-to-one, or one-to-many relationship. In a one-to-onerelationship between buyer and seller, both partiesare the primary customer or supplier for eachother. However, either the supplier or the custo-mer will dominate the relationship and the otherwill conform to the dominant’s standards. In amany-to-one or one-to-many relationships, onelarger supplier or customer deals with manyother smaller businesses or consumers, which allconform to the larger organization’s businesspractices.

Conversely, in a loosely coupled model, noparty has the power to dictate the system used toconduct transactions. In a loosely coupled envir-onment a company does not have one mainsupplier or customer, instead it may deal withwhichever company it pleases, taking into accountsuch things as price, value, and reputation. Inorder to complete the transaction, the companiesmust either constantly be changing their businesspractices to adapt to new partners or there must besome sort of mitigation system which can resolvedifferences between partners. This would allow abusiness to form a relationship with another withminimal effort or changes to their e-commerceinfrastructure. However, the forming of such amitigation system is not trivial and may requireserious investment.

In order to illustrate the concepts in the sub-sequent sections, let us consider the followingexample. Suppose an organization AUTO is anautomobile manufacture and it essentially assem-bles the parts together to assemble an automobile.Some of the parts such as the engine may bemanufactured in-house while other parts such astires may be purchased from other organizationssuch as PART1 and PART2. Suppose the con-sumers are CON1 and CON1. CON1 and CON2place orders to AUTO. AUTO then has to order itsparts from PART1 and PART2. These parts have toarrive in a timely manner. Then AUTO putstogether the parts and assembles the automobilesfor CON1 and CON2. In this case CON1 andCON2 are the consumers of AUTO. AUTO is theconsumer of PART1 and PART2. However, tocarry out the orders by CON1 and CON2, thethree organizations AUTO, PART1 and PART2have to collaborate with one another.

In certain cases, CON1 and CON2 may also haveto collaborate with AUTO to ensure that the partsare delivered according to the specifications. Weillustrate this example in Figure 8. In this example,one could have both loose coupling and tightercoupling models. For example, PART1 may betightly coupled with AUTO and this could meanthat PART1’s only customer is AUTO. PART2 maybe loosely coupled with AUTO. In this case, AUTOcannot dictate to PART2 and PART2 can have thefreedom to supply parts not only to AUTO but alsoto another manufacturer AUTO2.

Figure 8 C-commerce example



FEDERATED ARCHITECTURES FORC-COMMERCE

In a business-to-business c-commerce environ-ment, many corporations may have to collaboratewith each other (Shim et al., 2000). The variousc-commerce sites have to form federations so thatthe organizations can collaborate on transactionsand projects and yet maintain their autonomy.That is, a collection of cooperating c-commerceservers, which are possibly autonomous, may forma federation. The intent is for an organization tocontinue its local operation and at the same time

participate in a federation if it wants to. Archi-tecture for a federated environment to carry outc-commerce is illustrated in Figure 9. In thisarchitecture, the e-commerce sites are connectedthrough an EFDP (e-federated distributed proces-sor). The various e-commerce servers form afederation and have to cooperate with one another.They also have to maintain some kind of auto-nomy. The administrator of an e-commerce sitewould want to have as much autonomy aspossible to carry out intra-organizational opera-tions. At the same time he would want to carryout as many inter-organizational transactions as

Figure 9 Federated architecture for c-commerce



possible. That is, the administrators of the differentc-e-commerce sites would have to cooperate withone another to share each other’s data as well ascarry out transactions. Autonomy and cooperationare conflicting goals and therefore a balancebetween the two has to be achieved. Autonomy,which enables an organization to join or leave afederation whenever it wishes to, makes the taskof developing a c-commerce environment quitecomplex. While there is much research to be doneon federated architectures for e-commerce, manyof the concepts and techniques for federateddatabase management discussed in the paper bySheth and Larson (1990) can be applied here. Wediscuss some of the issues.

In federated databases, various aspects of het-erogeneity have been examined. These includepolicy heterogeneity, schema heterogeneity anddata model heterogeneity. All these aspects haveto be considered for federated e-commerce organi-zations. More importantly, each organization mayenforce a different business model. For example,one organization could enforce a model wherepayments are received within a month whileanother organization may enforce a model wherepayments are received during purchase. This isjust a simple example. There may be many otheraspects to heterogeneity with respect to businessmodels. Therefore, if organizations have to formfederations and carry out business-to-businesstransactions they have to ensure that the differ-ences are reconciled. Business model heterogeneityis illustrated in Figure 10. Essentially, each organ-ization may have different business rules and the

differences have to be reconciled. We have media-tors to handle representation schemas and brokersto make negotiations. Its should, however, be notedthat languages such as XML (eXtensible MarkupLanguage) are providing common representa-tion schemes for carrying out transactions on theweb. In addition, various frameworks are beingproposed for E-Commerce (see Rehfeldt andTurowski, 2000). These frameworks have to takefederations into consideration.

Schema and policy heterogeneity are also impor-tant aspects for federated organizations. Here, theissues may be similar to those for federateddatabases. However, a more extensive investiga-tion of the issues needs to be carried out. Figure 11illustrates Sheth and Larson’s view of handlingschema heterogeneity for e-commerce organiza-tions. With respect to policies, Figure 12 illustratessecurity policy integration for e-commerce organ-izations. We have essentially used Thuraisingham’sview of integrating security policies for federateddatabases (Thuraisingham, 1994). Note that in thecase of schemas, each organization may exportschemas, which may be a subset of the schemasfor the local organization. However, in the caseof security policies, the export policies may bemore restricted. That is additional access controlmechanisms may be needed for foreign organiza-tions to access data at another site. Policies are notrestricted to security. They also include integritypolicies and administrative policies. The differ-ences between organizations have to be handledfor all types of policies. The business rulesdiscussed earlier may also be a type of policy. For

Figure 10 Business model heterogeneity



further discussions on security we refer to Bertinoet al. (1999, in press).

Query processing and transaction managementhave also been studied for federated databases.When conducting business to business e-commerce

one party may need to query information from theother organization. This information is necessaryfor the parties to perform the transaction effi-ciently. Information-retrieval techniques have tobe integrated with security policies as well as

Figure 11 Schema integration and transformation in a federated environment

Figure 12 Policy integration and transformation



business rule processing. Transactions in business-to-business e-commerce are defined more broadlyand are essentially activities that are carried outacross organizations. They could be purchasinggoods, merging, as well as acquisitions. Whiletransactions for federated databases are involvedwith database transaction processing, the trans-actions for business-to-business transaction includedata transactions as discussed in Ghosh (2000) aswell as database transactions. We need socialtransaction models for federated c-commerceorganizations. There is research on transactionprocessing for e-commerce, but a lot remains tobe done. For example, how do you carry out con-current transactions? What happens if the trans-actions fail? There are not only technologicalconsiderations but also business considerations.

In the case of federations, while organizationscooperate and collaborate with each other, theyalso have autonomy. Sheth and Larson (1990)have discussed various aspects of autonomy. Oneparticular type of autonomy, very relevant tofederated organizations, is communication auto-nomy. For example, Organizations A and B maywant to collaborate with each other and carriedout business-to-business transactions and the samewith Organizations B and C. However, A and Cmay not want to carry out any transactions. Theprivacy of A and C have to be protected by Bwhen carrying out transactions.

Let us now examine our AUTO example. Eachof the organizations AUTO, PART1, PART2,CON1 and CON2 may have different schema,which they may export to federations. For exam-ple, PART1 may export only certain schema to

AUTO while it may export certain other schema tosay another manufacture which which we will callAUTO2. In addition, the different organizationsmay enforce different security policies. AUTO maygive access to all its automobiles to CON1, while itmay give access to only certain automobiles toCON2. Finally, the different organizations mayhave different business models. As mentionedearlier, AUTO may have a loosely coupled modelwith PART2 while it has a tightly coupled modelwith PART 1. Many of the techniques developedfor federated data management may have to beadapted for c-commerce.

ROLE OF KNOWLEDGE MANAGEMENT

The previous sections discussed models and archi-tectures for c-commerce. The question is, how doesknowledge management support c-commerce?Since training is part of knowledge management,knowledge management helps e-learning andeventually would help e-commerce as well asc-commerce. One needs to learn from experiencesexperts when creating and managing c-commercesites. One can also use previous experiences tocarry out transactions with organizations. Weillustrate this in Figure 13.

Finally what are the implications of know-ledge management and c-commerce for businessand management? Organizations want to be costeffective. While becoming leaner and meanerorganizations want to avoid duplication of effort.They want to reduce staff and employ knowledgeworkers. Knowledge management plays a strategic

Figure 13 Knowledge management for c-commerce



role here. Based on previous experience, organiza-tions can avoid serious mistakes as well as avoidcarrying out business with undesirable partners.For example, AUTO may have had adverseexperience with PART1. In this case AUTO maywant to stop its business association with PART1and join a federation with, say, another organiza-tion PART3. Knowledge management tools willalso make managers more efficient so that they canfocus on leading the organization instead of justmanaging it. We see knowledge management andc-commerce to be the central components of anyeffective and successful organization.

SUMMARY AND DIRECTIONS FORC-COMMERCE

This paper has described c-commerce. It is essen-tially integration of e-commerce, knowledge man-agement and collaboration technologies. We firstdiscussed the building blocks for c-commerce andthen described models and architectures. Then weaddressed the strategic role of knowledge manage-ment for c-commerce.

As noted in this paper, there are several areasfor further research. This paper has provided somedirections from models and architectures based onthe work carried out for federated databases. Whatare the suitable architectures for c-commerce?How can we handle differences with respect tobusiness models, schemas and polices for differentorganizations? What are the security concerns(Ghosh, 1998)? What information managementtechnologies are applicable? How do we developstandards for c-commerce? We believe that thereare plenty of opportunities for research intosuitable models, architectures, standards and tech-nologies for c-commerce. In addition, one also hasto consider the managerial and business implica-tions of c-commerce. Our ultimate goal is to buildorganizations that show strong leadership and arecost effective.

ACKNOWLEDGEMENTS

We thank P. Huck of MIT for his assistance with thispaper.

REFERENCES

Bertino E, Ferrari E, Atluri V. 1999. The specification andenforcement of authorization constraints in workflowmanagement systems. ACM Transactions on Informationand Systems Security 2(1).

Bertino E, Castano S, Ferrari E, Mesiti M. in press.Specifying and enforcing access control policies forXML document sources. World Wide Web Journal 3(3).

Caldwell NHM, Clarkson PJ, Rodgers PA, Huxor AP.2000. Web-based knowledge management for distrib-uted design. IEEE Intelligent Systems May–June: 40–47.

Chen S. 2000. Global leadership in c-commerce. Com-puter Software and Applications Conference, October,25–27.

Flynn L. 2000. Value creation: serving the digitalcustomer. Knowledge Management Magazine January.

Ghosh A. 1998. E-commerce security, weak links and strongdefences. John Wiley: New York.

Ghosh. 2000.Ginsburg M, Kambil A. 1999. A Web-based knowledge

management support system for document collec-tions. Proceedings of the 32nd Hawaii InternationalConference on System Sciences.

Malhotra Y. 2000. Knowledge management for e-business performance: advancing information strategyto ‘Internet time’. Information Strategy, the Executive’sJournal 16(4): Summer, 5–16.

Morey D, Maybury M, Thuraisingham B (eds). 2001.Knowledge Management. MIT Press: Cambridge, MA.

Rehfeldt M, Turowski K. 2000. Business models forcoordinating next generation enterprises. IEEE March.

Saunders VM. 2000. Collaborative enterprise environ-ments enterprise-wide decision support and know-ledge management. Aerospace Conference Proceedings(Vol. 2). IEEE, March, 321–330.

Scott J. 2001. Interogranizational learning and informa-tion technology in global manufacturing. GraduateSchool of Management — University of California,Irvine, Research in progress. Available at http://www.bus.utexas.edu/yjarvenpaa/scott.html

Sheth A, Larson J. 1990. Federated database systems.ACM Computing Surveys September.

Shim SSY, Pendyala VS, Sundaram M, Gao JZ. 2000.Business-to-business e-commerce frameworks. Com-puter October: 40–47.

Thuraisingham B. 1994. Security for federated databasesystems, computers and security.

Thuraisingham B. 2000. Web Information Management andElectronic Commerce. CRC Press: Boca Raton, FL.

Tracy KW. 2000. What’s the deal with ‘e-business’? IEEEPotentials February–March: 34–35.

Winch G. 2000. Knowledge management and competi-tive manufacturing. Engineering Management JournalJune: 130–134.



collaborative commerce and knowledge management

Documents