Best Practices in Enterprise Knowledge Management

KMWorldSupplement to

Premium Sponsors

November/December 2004

Andy Moore . . . . . . . . . . . . . . . . . . 2 Why We Were Wrong About KM …and How to Make It RightI just heard a song on the radio. The lyric in the chorus is: “Shout all you want, itdoesn’t matter; 70% of me is water; 70% don’t care about the rent; the rest ofme is finding it harder…” Thanks to that ridiculous lyric, I’m wondering if wearen’t all wrong about knowledge management. I am now wondering whether 70%of any organization is water. Or, in other words, are our organizations largely madeup of an inert element whose purpose is simple and mundane. . . ?

Jay Weir, Hummingbird . . . . . . . . . . 4 Bringing Structure to the Unstructured—Total KnowledgeManagement

Winning in the competitive global marketplace requires agility in order to leveragecontent dispersed across the enterprise, enhance service delivery, improve efficiencyand reduce risk. . . .

Alkis Papadopoullos, Convera . . . . . . 6 Information Analysis and the Content Supply ChainAccess to content in many different formats has been growing in importance as companies and government institutions attempt to get a better handle on the information at their disposal. . . .

Christina Robbins, Digitech Systems. . . . 8 A Single Solution—Cost Savings and ComplianceThought HIPAA compliance would cost you an arm and a leg? It doesn’t have to. There are scalable solutions that will convert your paper-based records to electronic files. . . .

Lee Phillips & Bjorn Olstad . . . . . . . . . 10 Winds of Change—Extracting Meaning from ContentOver the last four years, the Internet has changed many tried-and-true businesspropositions. From category-killers like Amazon, to new-paradigms like eBay and Priceline, to new-media giants Google and Yahoo!. . . .

Tom Tobin, ServiceWare . . . . . . . . . 11 The Top 5 Reasons You Need KM—TodayThere are undoubtedly many projects on your corporate wish list, all tagged as“high priority.” Many of the projects sound good, some seem interesting and they all promise incredibly fast ROI. . . .

Max Schireson, Mark Logic . . . . . . . 12 Does Technology Matter for Knowledge Management?When knowledge management implementations fail, the blame usually falls onchange management, business process or some other soft factor. After all, the proximate cause of the failure is most typically. . . .

Steve Papa, Endeca . . . . . . . . . . . . 13 A Primer on Faceted Navigation and Guided NavigationYou've likely heard the buzz by now: faceted navigation truly changes the groundrules for KM, and for the many related applications where users need to find information, ranging from subscription content services to directories. . . .

Noetix . . . . . . . . . . . . . . . . . . . . . 14 Maximize Your Investment—Empowering CRM Users withKnowledge

In accordance with your overall KM strategy, your organization has decided toinvest in a customer relationship management (CRM) system. After a thorough vendor analysis, Siebel is chosen. . . .

Fast Search & Transfer

attacked it from different directions, andbrought in various systems to organize andmanage some aspect of information. Butthey couldn’t deal with all of it.”

Because, perhaps, they didn’t need to.The low-hanging fruit for any KM systemvendor has always been the organization’srelatively small legal department, Saundersnotes. And rightly so; there is a self-evidentadvantage in making sure all the lawyersare aware of all the risks and opportunitiesat play at any given time, enterprise-wide.That information is found only in theunstructured document-set.

Saunders points out that more recently,the CRM function of a company has alsobegun to benefit from knowledge inter-change. It's also limited to a fairly smalland specialized group, she points out, but

managing customer information can besaid to have great value there, too.

But the same value proposition can’t bemade as easily for a financial servicesorganization, and even less so for a trans-portation/shipping company or a retailchain. Their information value propositionis more skewed toward structured data—earnings reports, inventory status from theERP system, logistic information from ship-ping and receiving manifests. Anecdotesand story-telling may be great for lawyers,but not so much the guy shipping containersacross the North Sea.

“Making the entire organization ‘knowl-edge-aware’ might be an unrealistic goal,”says John Bellegarde, Hummingbird’s vicepresident product management. “Some peo-ple in an organization don’t benefit fromknowledge management in their day-to-daylives. It isn’t a cure-all to all problems. It’show much an individual ‘trades’ in theirknowledge, and how much their expertise isdependent to their job function, whether ornot having more readily available KM willbenefit them.”

Granted, as Mathias Evin(Hummingbird’s product manager for dataintegration solutions) points out, structureddata management and consolidation is aproblem pretty much solved. There is muchbuzz—largely coming from the vendor com-munity—to find ways to tie structured reportswith unstructured information to provide acomplete picture. The degree of traction thismovement will gain remains to be seen.

But even if the two worlds unite in aseamless partnership, there will be only asmall portion of any organization that willbenefit from such a united knowledge-base.That 70% will likely have no use in theirday-to-day lives for such a holistic view.

Truly knowledge-intensive industriesmight be the exception. Large law firms, forexample, and professional service organiza-tions have “knowledge teams” (I’d love tohear their cheers) that are focused not on thecompany’s mission-critical processes, but ondeveloping the knowledge bases, addingvalue and improving the gathering process.It’s another level of abstraction for a compa-ny like that—knowledge IS the product intheir cases. But for most of the world, knowl-edge work is fairly narrowly limited to asmall group. I happen to think it’s that 30%from that darn song lyric.

So, what’s the best practice for bringingknowledge management to your organiza-tions? Pick your fights. “KM is never suc-cessfully deployed as a huge enterprise-wide effort,” Saunders points out. “Usually,

Supplement to

Why We Were WrongAbout KM...and How to Make It Right

I just heard a song on the radio. The lyricin the chorus is:

“Shout all you want, it doesn’t matter;70% of me is water;

70% don’t care about the rent; the restof me is finding it harder.”

That ridiculous lyric is now hopelesslystuck in my brain, and somehow leading meto think about knowledge management, ofall things.

I’m wondering if we all weren’t wrongabout knowledge management. I am nowwondering whether 70% of an organizationis water. Or, in other words, are organiza-tions mostly made up of an inert elementwhose purpose is simple and mundane—while the 30% of our companies are doingall the hard work?

Despite the Carrie Bradshaw-like toneof that last sentence, I mean it seriously.The knowledge management industry (ifyou can call it that) has always taken apretty high-falutin’ attitude about “theenterprise,” and it’s possible that its well-documented image problems stem fromthat strategic flaw.

KM has long been touted as an all-or-nothing proposition. It is believed that thesuccess of any worth-its-salt “knowledge-based enterprise” is directly proportionateto the extent to which every single memberof the organization can interact with eachother and the company’s various “knowl-edge repositories.” From the janitor to thechairman, a knowledge-based organizationhas to be utterly permeable to be agile.

Well, pshaw. I have a different idea.The evidence is mounting that the

famous “departmental point solutions” thatmany of us thought were mere baby steps toa greater goal, are, in fact, the end game. “

“The market has changed quite a bit fromwhen KM made headlines in the ‘90s, whenit was thought to the magic bullet,” agreesMarcelline Saunders, who is in charge ofknowledge management at Hummingbird.“They realized there was value in the infor-mation people possessed in their heads; theyjust didn’t know how to get at it. So they

November/December 2004S2 KMWorld

Andy Moore has heldsenior editorial andpublishing positions formore than 25 years. As atechnology writer andeditor, Moore speaks withdozens of seniorexecutives and industryexperts each month. Inhis role as EditorialDirector for the SpecialtyPublishing Group, Mooreoversees the

contributions to the series as well as conducting marketresearch for future topics of interest for the series.

Moore was the editor-in-chief of KMWorld Magazine andis now its publisher

Andy Moore

By Andy Moore, Editorial Director, KMWorld Specialty Publishing Group

“KM has long been

touted as an all-or-nothing

proposition. Maybe, just

maybe, it’s not.”

Supplement to November/December 2004 S3

a small project is identified where there isdefinitely an ROI. There’s got to be a bene-fit that can be identified quickly,” she says,and it may be as small as avoiding mailingcosts by deploying an electronic documentmanagement solution.

As for executive buy-in (which we havealways identified as a key component),there still has to be that kind of mandate.BUT, the “good” executive decision is onethat locates those specific areas where thereis immediate benefit, and in those rare butbeautiful cases where departments need tocommunicate between silos and through thefirewalls.

Just be careful to recognize: KM doesn’tsolve every problem, and shouldn’t even try.

Market View: Consolidated orSpecialized?

Mark Myers, vice president of strategicalliances at Convera, is talking: “You canrecap the tops of the waves of KM for thepast 15 years or so this way: First, you hadthe ‘applications’ that vendors sold as KMsolutions. Then you had a portal craze, wherepulling everything into a single view meantyou were managing it. Now, the key move-ment is consolidation...and who knowswhere that will go....enterprise suites, I sup-pose ...” (not sounding too convinced).

I’ve known Myers for a long time, and onthe day we talked, he and I were both kindatired. Well, I was. Myers had been at a semi-nar in Houston, co-sponsored by energy-exploration giant Andarko, and was upbeat,as usual, and willing to talk about the state ofthe marketplace in a free-wheeling conversa-tion that touched on the enterprise market,the search business and the opposing forcesof consolidation versus specialization.

“If you have a 30,000-person, globalorganization and you want to manage allyour knowledge, you’re going to be talkingto Gartner and the rest about an enterprise-wide suite of products of some kind. But ifyou look at the Magic Quadrant, you’ll real-ize that it is not quite yet a mature market.And you might then turn to IBM orMicrosoft or Sun, depending on your tech-nology orientation,” said Myers.

“So we’re definitely beyond the three-legged stool of ‘people, process and technol-ogy’ when we talk about knowledge manage-ment at that level,” according to Myers.Strategically, he feels, companies are viewingKM through the filter that matches theirneeds most appropriately, with the expecta-tion that these various “interest domains” allfeed-up into, eventually, an enterprise solu-tion of some kind. But it’s not there yet.

“You may be gravitating around content,so your focus is on ECM,” Myers explained.“Or maybe it’s collaboration, so you’relooking more closely at messaging and dis-

cussion groups. Or it might be search andcategorization, or business processes...how-ever you’re entering the stream, there’s stilla long way to go before you’ll have the solu-tion you’re looking for. There’s a lot of workstill to be done in R&D, etc. So these areasare not tapped out.”

He went on to bring it into more “insidebaseball” terms: “The industry macroeco-nomic question is whether it will all getconsolidated into a few big offerings, andwithin those offerings these ‘componentareas’ will get fleshed out, or will we havea gradually declining number of independ-ent vendors doing cutting-edge work?”

What, I asked, does it look like fromyour industry’s perspective?

“Well, it’s a great example,” he said“There have been acquisitions in the searchbusiness, but there are still four or fiveviable vendors competing for the enterprisebusiness. In the other areas, I think you’llfind pretty much the same thing.”

So the interesting conundrum is this:While best-of-breed solutions are stillactively serving the marketplace, is it safer(for your overall KM strategy as well as foryour career) to adopt the “one-stop-shop”concept of the enterprise suite vendors?(By the way, with Oracle apparently enter-ing the picture in an apparently seriousway, this decision has picked up a littlemore gravitas in recent weeks. Watch thisspace for more on THAT subject...)

Well, the answer to the conundrum, asusual: is: It depends. “For any of the areascovered by the enterprise suites, there is aclass of customer who will buy best-of breed,”explained Myers “The search that’s goodenough for the transportation sector isn’t goodenough for a pharma company. There’s toomuch at stake for pharma companies to sim-ply accept the ‘packaged’ search tool.”

The companies who need to be on theedge and require advanced search (to take anexample, but the same is true of BPM andcollaboration and all the rest) are themselvesagile and engaged enough to derive businessvalue from their decisions, Myers believes.

So how do organizations decide whetherto go “best of breed” or not? In many cases,the decision is practically made for them,through competitive or other stimuli. Forexample, Myers points to government. “Thegovernment agencies that regulate industriesthat are themselves very knowledge-drivenneed to simulate the type of environment

those industries have created. So the FDA isusing the same level of tools that pharmaceu-tical companies use. One is doing the R&D;the other is analyzing what they did.”

Following on from the conversationwith Hummingbird earlier, it is Myers’contention that even within organizations,there remains a mix of solutions, and thosehave been chosen to address specializedneeds—sometimes rightly, sometimeswrongly (“Sometimes there’s one guy whojust becomes convinced one product is bet-ter than another,” Myers says.)

“Take ECM, for example. If you scratchhard enough, you’ll see that there is noTRUE enterprise content management.There are products whose aspiration is tomanage every bit of content for the company.But it just doesn’t happen. It happens to us:the FBI and the FDA use our product, but Iknow there are other search-software licens-es in those organizations,” Myers contended.

“A pure enterprise deal doesn’t happenfor any vendor, anywhere.”

So the bottom line is that choosing solu-tions for your organization is not easy. Butit’s possible, thanks in equal parts to theemerging enterprise suite vendors who arereaching their crucial mass, as well as thespecialized vendors who target specific andunique sets of needs. How do you choose?Well, you start here. I can’t think of a bettergraduate-level course in information man-agement than reading and learning from themany authorities who have shared theirexpertise in these KMWorld White Papers.This one is certainly no exception. ❚

Andy Moore is a 25-year publishing professional, editor andwriter who concentrates on business process improvement throughdocument and content management. As a publication editor, Mooremost recently was editor-in-chief and co-publisher of KMWorldMagazine. He is now publisher of KMWorld Magazine and its relatedonline publications.

As Editorial Director for the Specialty Publishing Group, Mooreacts as chair for the “KMWorld Best Practices White Papers” and the“EContent Leadership” series, overseeing editorial content, conductingmarket research and writing the opening essays for each of the whitepapers in the series.

Moore has been fortunate enough to cover emerging areas ofapplied technology for much of his career, ranging from telecom andnetworking through to information management. In this role, he hasbeen pleased to witness first-hand the decade’s most significant busi-ness and organizational revolution:the drive to leverage organization-al knowledge assets (documents, records, information and objectrepositories) to improve performance and improve lives.

Moore is based in Camden, Maine, and can be reached atandy_moore@verizon.net.

KMWorld

“We’re beyond the three-legged stool of

‘people, process and technology.’”

and other stored documents. This 80% ofunstructured information is often over-looked as a key information source for tac-tical and strategic decision making. Searchengines and classification technologieshave been the tools of choice for accessingunstructured information, but their relatedmetadata is not used the same way as struc-tured resources.

Many organizations choose to manageeach stage of the content lifecycle independ-ently, with disparate processes, systems,repositories and technologies. Today, withdirectives to increase efficiencies whilesimultaneously improve accountability, anintegrated system for the management of theentire content lifecycle is crucial.

At the most basic level, one of the cur-rent business drivers in information man-agement is to improve organizational effi-ciency and reduce costs. The next step in theprogression is to increase business value,including that which knowledge manage-ment yields from an organization’s informa-tion stores. Lastly, once these goals areachieved, organizations move to increasetheir focus on managing risk with elementssuch as records management initiatives.This includes managing content assets tokeep what is important or mandatory forlegal, compliance, historical and knowledgemanagement needs, and retiring or dispos-ing of information no longer useful or rele-vant as its value has expired.

Organizations need to leverage technolo-gy to meet business continuity and corporategovernance requirements, for making busi-ness content traceable secure and auditable,at the same time they invest in aligning busi-ness processes and content to optimize per-formance and reduce risk exposure.

Today, many organizations are manag-ing each stage of the content lifecycle inde-pendently, with disparate processes, sys-tems, repositories and technologies. Thisresults in the following issues:

◆ Incomplete, inaccurate and limited infor-mation;

◆ Time wasted locating the right information;◆ Poor collaboration between groups;

◆ Difficulty getting fast access to the fullpicture;

◆ Maintenance of disparate systems;◆ Difficulty keeping systems consistent;◆ Difficulty delivering the right information

to the right people;◆ Poor asset management—information is

not used to its full potential; and◆ Weak compliance—compliancy with reg-

ulatory rules is difficult to ensure.

Making Total Knowledge Managementa Strategic Advantage

Content assets are the foundation uponwhich corporations build new businessopportunities and competitive advantage—and upon which governments build account-ability practices and ways to better servetheir constituents. Leveraging the businessvalue of content while simultaneously pro-tecting that content with an applied recordsmanagement strategy is not a new concept;however, the conventional means of address-ing these tandem needs is changing.

Today, total knowledge management isdriven by the need for organizations tohave a unified access and view to all busi-ness content across an enterprise. Businesscontent needs to be linked, referenced andintegrated to get a complete picture. Thisbrings context to all content beingaccessed, searched, reported on, collabo-rated on and analyzed.

The Total Knowledge Management Architecture

A total knowledge management archi-tecture is a blueprint that provides theessential functionality to manage informa-tion and knowledge though its lifecycle,from inception to destruction, whileenabling organizations to use assets to theirfull potential and maximize their return oninvestment.

Getting at the mountains of raw datagenerated by most modern organizationsfrom a variety of different hardware plat-forms, systems and applications is a diffi-cult task. Organizations that want to gain akey competitive advantage need to thor-oughly implement a comprehensive solu-tion that will help them manage their con-tent asset lifecycle. This means that deci-sion makers are able to efficiently andsecurely get access to any structured andunstructured information they require atthe time they need it.

A single interface allows users toaccess, search, report, collaborate, analyze,track and audit against both structured andunstructured enterprise content. To ensurea thorough lifecycle management of thisasset, organizations need to implement a

Supplement to

Bringing Structure tothe UnstructuredTotal Knowledge Management

Winning in the competitive global mar-ketplace requires agility in order to leveragecontent dispersed across the enterprise, en-hance service delivery, improve efficiencyand reduce risk. Managing enterprise con-tent within an organization and using it to itsfull potential is critical to success. This meansmanaging both “structured” and “unstruc-tured” content, business processes, line-of-business solutions, all while maintaining thelowest total cost of ownership (TCO).

The ability to systematically capture, cre-ate, manage, access, review, distribute, pub-lish, store and preserve all business con-tent—from ERP/CRM systems, databases,e-mails, documents, file systems and externalinformation systems—and leverage thisinformation to its full potential represents theconcept of Total Knowledge Management.

We refer to total knowledge as havingcomplete insight gained through 100% uti-lization of “organizational informationassets”—both structured and unstructuredbusiness content across an enterprise.

What is Driving Total Knowledge Management?

Today, fierce competition is forcing com-panies to be smarter—to leverage all enter-prise content assets, uncover valuableinformation hidden in multiple systems andfrom structured and unstructured sources.

In the past, reporting on and analyzingstructured information was only utilized tounderstand the business environment, cus-tomers, suppliers and other key businessviews. Tactical and strategic decisionswere made based on structured informa-tion found in relational databases, legacydata stores and various reporting systems.Structured information only represents20% of a company’s enterprise informa-tion assets. The remaining 80% of valuableinformation is in uncover form dispersedacross the enterprise—in e-mail, documentmanagement systems, file systems, instantmessaging systems, records managementsystems, knowledge management systems

November/December 2004S4 KMWorld

By Jay Weir, Senior Manager, Product Marketing,Hummingbird Ltd.

Supplement to November/December 2004 S5

comprehensive solution that helps themmanaging those multiple stages.

◆ Business process management andworkflow touches the entire enterprisecontent lifecycle. The first aspect of thetotal knowledge management solution isto implement a business process man-agement (BPM) element. BPM ensuresautomated and controlled process im-plementation, integration of various enterprise content components, captur-ing of the different events of the enter-prise content lifecycle and streamlinedmaintenance.

◆ Data management is the second aspectof this comprehensive solution. Organi-zations need to distill, consolidate, or-ganize and classify huge amounts of dataand dynamically transform it into accu-rate, consistent and timely information.During this process, this vast amount ofstructured data and unstructured data isenriched and gradually converted into asmaller amount of meaningful content.This corresponds to the data managementaspect of this comprehensive solution.This data management aspect encom-passes document management, data in-tegration, knowledge management, col-laboration, record management andreporting.

◆ Knowledge management and informa-tion retrieval technologies are critical tomanaging unstructured as well as struc-tured information. It is imperative tocapture best practices, expertise and in-formation. Proper tools enable knowl-edge workers to find information that isrelevant and accurate for better decisionmaking.

◆ Metadata management is the final ele-ment of the strategy. During different

stages of the content lifecycle, a great dealof metadata (information about informa-tion in its simplest form) is created. Thismetadata brings additional meaning to theinformation, and turns information into“content in context”. Therefore, thismetadata needs to be captured, managed,reused and deployed to get the full mean-ing of underlying information. Metadatamanagement provides for metadata ex-change, data lineage, impact analysis,data linkage, content documentation and,perhaps most importantly, supports andstreamlines regulatory compliance ef-forts. Metadata is very important in searchand classification where it plays a role indetermining the relevancy of informationand helps to narrow a search (i.e. facetedsearch). Metadata is also used in the creation of non-subject taxonomies. Unstructured information often has metadata within the document that needsto be identified and extracted in order tobe used as well as associated metadatalinked to it in other repositories (CRM,document management, etc.).

These aspects form a comprehensive“total knowledge solution” and are essentialto ensure accurate and timely access toenterprise content.

Where Do You Start?Building a total knowledge manage-

ment infrastructure is not as simple asimplementing a set of technologies thatinstantaneously allow a company to lever-age all their information assets. Creating thenecessary capabilities to meet a company’sneeds requires a combination of people,process and technology and must be per-formed in line with the company’s overall

strategies, goals and objectives. The quick-est way to failure is to try and implementnew technologies into an existing businessprocess. This typically results in inefficien-cies that cause users delays in accomplish-ing their jobs. Having a total knowledgeview of the business should be viewed as aseries of business processes and userresponsibilities that when followed willenable the organization to meet theirresponsibilities.

Hummingbird Enterprise is designedaround the principle that enterprise contentmanagement functionality must be trans-parently accessible from a wide range ofuser interfaces, the Web top, the desktop,e-mail clients, mobile devices and thirdparty portal frameworks.

◆ Reusable templates that support line-of-business enterprise content managementsolutions such as contract management,correspondence management and com-pliance;

◆ Process-centric workflow that persistsacross all the components within Hum-mingbird Enterprise;

◆ An integrated environment for storingand managing all types of business content, including documents, records,images and digital assets, e-mail, Webcontent, instant messaging threads andstructured reports;

◆ Bridging of structured and unstructuredbusiness content;

◆ Comprehensive e-mail management ca-pabilities including rules-based e-mailclassification; and

◆ Extensibility and interoperability be-tween Hummingbird Enterprise andthird-party applications.

The ultimate goal of the total knowl-edge strategy is to enable an organizationto have one single version of the truth; uni-fied access to all enterprise content; bettercompliance with regulatory rules; cost andproductivity losses reduction; and optimumusage of content assets. ❚

Jay Weir manages product marketing strategy on a global basisfor Hummingbird.With more than 10 years experience in informationtechnology management,Weir writes and speaks extensively on

Enterprise Content Management (ECM), knowl-edge management, and business intelligence.His knowledge areas range from data transfor-mation and replication,data mining, and business analytics to information search andretrieval technologies, portals, document andrecords management, and team collaboration,and he has been published in major industry

trade press, journals, and text books.

Hummingbird Ltd. is a leading global provider of enterprise software solutions, employing over 1,450 people in 40 officesworldwide. Hummingbird Enterprise™ 2004 is a state-of-the-artintegrated enterprise content management platform that enablesorganizations to securely access and manage business informationsuch as documents, records, e-mail or financial data. Pleasevisit:www.hummingbird.com.

KMWorld

Relationship between Structured and Unstructured Business Content

Jay Weir

must be accurate or accompanied with arelevance measure that allows the decision-maker to assess its potential usefulness orcredibility.

This brings us to what we believe to be abasic tenet of knowledge management,namely that the primary difference betweeninformation and knowledge is relevance.This, above all else, leads organizations torequire an information workflow solution.

Traditionally, the chain of informationmanagement has been broken down alongthe lines defined by the three prevailingtechnologies in the chain: portals, informa-tion analysis systems and search engines.Portals have served to accelerate the trans-formation of information into businessintelligence because they act as an inter-vening lens between search engines thatlocate macro data and information analysistools that parse micro data.

But, more and more, the lines betweenthese three key application areas havebecome blurred. Today, streamlining theinformation supply chain is no longer amere question of connecting to the numer-ous available repositories, and then makingthem available through the sequential fil-tering of search, portal and informationanalysis tools.

Portals include search capabilities andsome of the typical information analysistools that revolve around gathering andorganizing data. They also include toolsthat assist with dissemination, e-learning,publishing and data lifecycle management.Search and categorization tools are alsogrowing in functionality, not only taking onthe brute force task of retrieving indexeddata, but also of helping to pre-organize thatdata in such a way that the most relevantresults are presented in context of theimmediate information analysis needs ofthe person sifting through the data. Becauseof this functional consolidation, as well asorganizational restructuring and consolida-tion, the path to actionable knowledge—that is expected to be more readily availableand more readily quantifiable from an ROIperspective—lies in an integrated approachto managing the information supply chainthat we now coin as the “InformationAnalysis Portal.”

Information analysis portals are growingin importance and will likely be ubiquitous asorganizations move more rapidly to bettermanage their information. The urgency tomore efficiently and expertly process storesof information is being felt more and more.Businesses are seeking more effective waysof defining and developing significant prod-uct differentiators. Government agencies areseeking to more effectively and more reliablyassess national security threats, while afford-ing citizens greater access and control overinformation to help them avoid heavy andlengthy bureaucratic systems when seekinganswers to questions. This is evidenced in thedesire of organizations to consolidate theirsearch capabilities around a single platformthat affords the most control over several keyaspects of the information supply chain asmuch as possible. The chain has grown inimportance and complexity, and now encom-passes many varied features such as:

◆ Connectors for accessing numerous dataformats and media;

◆ Annotation tools for adding metadata,and linking data sets;

◆ Traditional searching capabilities for ac-cessing this content;

◆ Profiling and agent-based capabilities totrack key changes to information stores;

◆ Categorization and classification of datato best manage very large lists of relevantdata;

Supplement to

Information Analysis andthe Content Supply Chain

Access to content in many different for-mats has been growing in importance ascompanies and government institutions at-tempt to get a better handle on the informa-tion at their disposal. A significant portionof IT investments in the last two years hasgone into applications that extend beyondmore traditional CRM, ERP and databaseapplications to reach toward a new class ofcontent management, document manage-ment, collaboration and search and catego-rization solutions.

This push really has to do with manag-ing what we might call the “informationsupply chain.” In so doing, information(input) comes from all potential sources, asinformation can lie anywhere and mustyield knowledge (output) that is actionableand comprehensive. To achieve this, theknowledge output must be accessible intimely fashion, and, more importantly,

November/December 2004S6 KMWorld

By Alkis Papadopoullos, Director, Linguistic Technologies, Convera Canada

Search engines find information; portals tune the data noise until it becomes a useful signal; information analysis transforms that signal intobusiness intelligence that engenders profit and success.

Supplement to November/December 2004 S7

◆ Data translation for handling multiple en-codings and languages;

◆ Natural language processing to betterhandle a wider range of users and queries;

◆ Entity extraction as a basis for link andtemporal analysis;

◆ Entity relationship extraction;◆ Summarization of data to minimize the

need to parse whole data sets;◆ Better integrated processing of structured

and full-text data stores to reconcile tradi-tional data mining from databases and con-ceptual extraction from full-text sources;

◆ Geospatial analysis;◆ Collaboration for sharing pertinent data sets;◆ Expert location and communities-of-in-

terest; and◆ Reporting tools that serve to better enhance

functionality of analysis tools.

These can be grouped into overarchingcategories that represent the key processingblocks of the information supply chain.

Getting at the data.

Data Capture—the first knowledge management process known as gathering

This includes the connectors and theannotation tools mentioned above. Tappinginto all forms of file formats and media typesis crucial, but it must be possible in multiplelanguages as well. In addition, more toolsand functionality in widespread applicationssuch office suites allow for annotation ofdocuments, which are marked-up with keymetadata and even reference other docu-ments for further reading or relevant links.These should be exploitable by the dataaccess bridges that connect to all the variousdata sets.

Preparing the data for better analysis.

Data Normalization and Transformation—the second knowledge management processknown as organizing

This involves the ability to index contentwith linguistic markers such as languagetags, conceptual identifiers, grammaticalcategories (i.e., determining if a term is anoun, a verb, etc.) that allow for more accu-rate retrieval of results. It also involves theability to extract relevant entities, categorizecontent for subsequent classification andcompare against typical data profiles that

trigger events that serve to update usersabout changes in data. Of course, it meansbeing able to do all of this regardless of theencoding, the language or the quality of themedia type (as in the case of audio or video,for example).

Applying more sophisticated algorithms toextract valuable links in data.

Data Analysis—the third knowledgemanagement process known as refining

Data analysis revolves around naturallanguage processing of queries so as to beable to intelligently parse out a user’srequest even when written in so-called“natural language” (for example, a com-mon sentence rather than a complexBoolean query). It also involves the abilityto dynamically classify search results alongseveral dimensions, each dimension beinga branch of a taxonomy (or taxonomies).As an example, one could cross-referenceinformation across a geography branch andan infectious diseases branch if looking forgeographic distribution of a specific set ofdiseases. Additional analysis capabilitiesinvolve link, geospatial and temporalanalysis, which depend on, among otherthings, good entity relationship mappingand usage of such maps in conjunctionwith anaphora analysis, syntactic analysisand categorization algorithms.

Presenting the results to the user.

Data Reporting—the fourth knowledgemanagement process known asdisseminating

This fourth layer is where the end-useractually sees the fruits of the combinedfeature-functions being applied and wherethe information analysis portal’s key value

is best showcased. It involves the display-ing of information extracted through link,geospatial and temporal analysis via over-lays, maps and imagery. This is where col-laboration is significantly bolstered by theaddition of interfaces with other usersthrough on-line tools serving to bettershare information, whiteboard, co-produceand form communities of common inter-est. Within this context, all members bene-fit from profiling and agent-based alertsthat have been set up by individualsaccording to their unique expertise. It isalso where reporting and disseminationcan take on their full value through suchtasks as the creation of reports and analy-sis that are delivered in a timely fashion toa customer, a sales representative or a gov-ernment analyst.

The Bottom LineEffective information analysis depends

on comprehensive management of allaspects of the information or content sup-ply chain. The most effective way toaccomplish this is through the deploymentof an information analysis portal based ona fully integrated information discoveryand analysis platform. This portal shouldtypically include a good portion of theaforementioned capabilities and allow forcustom components to be added to the mixin order to ensure that all required func-tionality is available. This will help enablecompanies and government agencies aliketo address all key processes of informationanalysis and shorten the path to actionablecontent and profitable content managementsolutions. ❚

Alkis Papadopoullos,Director of Linguistic Technologies at Convera,directs the evolution of Convera’s language analysis, taxonomy develop-ment and discovery products—key components in Convera’sRetrievalWare 8 information discovery and analysis platform. He has aMasters degree in Physics,speaks five languages fluently and has workedin computational linguistics software development for ten years.

Convera is a leading provider of enterprise search and categoriza-tion solutions. RetrievalWare 8, Convera’s information discovery andanalysis platform, enables fast, personalized searches resulting in excep-tionally high productivity and returns on investment assets. For over 20 years, Convera has provided sophisticated search and categorizationsolutions to over 900 customers in 33 countries to power a broad rangeof mission-critical applications. For more information please visit www.convera.com.

KMWorld

“Streamlining the information supply chain is

no longer a mere question of connecting to

the numerous available repositories.”

“Information analysis portals

are growing in importance and

will likely be ubiquitous.”

Supplement to November/December 2004S8 KMWorld

designed to encourage adoption of secure,effective electronic document and contentmanagement practices in the healthcareindustry. As these systems are implemented,healthcare providers will realize tremendousbenefits including all of the following:◆ Reduced costs associated with filing, stor-

ing and retrieving patient information;◆ Improved efficiency, as documents can be

automatically routed through any organi-zational process including admissions,billing, collections and interactions withinsurance providers;

◆ Enhanced security with electronic audittracking—no more paper documentswhich are susceptible to misplacementand loss;

◆ Protected patient privacy resulting in im-proved patient confidence;

◆ Speedier “Explanation of Benefits” (EOB)and collections processes through real-timeaccess to electronic records and information,providing significant improvement in cashflow through reduced days payable; and

◆ Avoiding fines and litigation associatedwith noncompliance. In addition to the benefits that come from

HIPAA compliance, the healthcare providercan realize tremendous cost savings, stream-line organizational processes, and improvedocument security and storage through theuse of an effective document and contentmanagement system. Clearly, there istremendous motivation to adopt documentand content management, and the pressure ison as the April 2005 deadline looms. In fact,AHANews indicates that improving securityto comply with HIPAA is one of healthcareorganization CIOs’ top two priorities in2004. (“CIOs rate security, patient safety astop priorities for ’04, HIMSS survey finds,”Feb. 23, 2004.)

Selecting the Best SolutionWith all of this potential, healthcare

providers must take care in selecting the bestelectronic document and content manage-ment solution for their organization. Five keyareas of concern include disclosure tracking,protected disclosure, de-identification ofprotected information, patient disclosure

authorizations and generally protecting thesecurity and privacy of personally identifi-able electronic health information as it routesthrough the processes of diagnosing, treat-ing, reporting and billing individuals forhealth-related issues.

Disclosure tracking—HIPAA regula-tions require that the healthcare providertrack disclosures of personally identifiablehealth information to any internal or thirdparty. This includes doctors, insuranceproviders, billing and claims processingand all other entities that might have reasonto view the information.

Protected disclosure—The Privacy Ruleallows for certain types of disclosures that donot require authorization from the patient.These include: (1) to the individual; (2) treat-ment, payment and healthcare operations;(3) disclosures with the opportunity toobject; (4) incidental use; (5) public interestand benefit activities; and (6) release of alimited data set. There is specific informa-tion that must be captured for each disclo-sure, and, in some situations, only a subset ofthe patient’s record may be disclosed.

De-identification—There are no restric-tions on the use of de-identified health infor-mation, and much of this type of informa-tion is used by the government, researchersand health organizations in compiling statis-tics. De-identification requires the user tostrip all personally identifiable information(i.e., patient name, identity number, address,etc.) from the health document before it ispassed on.

Patient authorization—A healthcareprovider may not disclose protected healthinformation (PHI) outside the scope of theHIPAA regulations without written authori-zation from the patient. This authorizationmust be obtained, recorded and then main-tained for a period of at least six years. Ingeneral, authorizations should be specific asto the information being disclosed, the per-son disclosing and receiving the informa-tion, expiration, etc., so a separate authori-zation is often required for each disclosure.

Security—The Security Rule portion ofthe HIPAA regulations applies to the trans-mission of electronic protected health infor-mation (EPHI). The healthcare provider is toensure the confidentiality, integrity andavailability of all EPHI the covered entitycreates, receives, maintains or transmits. ❚

Christina Robbins is the Marketing Program Manager forDigitech Systems. She has a diverse background in software development and telecommunications,and joined Digitech in January2004. Christina develops marketing and communications programsdesigned to strengthen the effectiveness of Digitech’s reseller base.

A Single SolutionCost Savings and Compliance

Thought HIPAA compliance would cost youan arm and a leg? It doesn’t have to. There arescalable solutions that will convert your pa-per-based records to electronic files, storethose files in an ultra-secure system and al-low you to control the distribution of pro-tected information, thereby ensuring the se-curity and confidentiality of your sensitiveinformation.

The Impact of HIPAAWhen Congress passed the Health

Insurance Portability and AccountabilityAct (HIPAA) in 1996, it set forth variousprovisions for the privacy of protectedhealth information (PHI) during diagnosis,treatment and billing for medical issues.As a result of this regulation, the U.S.Department of Health and Human Services(HHS) issued the “Privacy Rule” to guidehealthcare providers through the imple-mentation of HIPAA-compliant processesand procedures, and the “Security Rule,”which extends those protections to elec-tronic information.

First Consulting Group, in a study prepared for the American HospitalAssociation, estimates that the healthcareindustry will spend up to $22.5 billion onelectronic document and content manage-ment solutions between 2001-2005. As clar-ifications continue to emerge regardingHIPAA, the HHS Privacy Rule and the HHSSecurity Rule, experts expect that spendingwill continue to increase.

ARMA International (arma.org) hasstated, “Most healthcare organizations mustcomply with HIPAA’s Privacy Rule by April14, 2003, but many other organizations—including a large number of employers—also will be affected by this rule. In fact,HIPAA’s Privacy Rule will impact, at leastindirectly, all organizations in some way.”Whether through increased security on elec-tronic transmission of health-related infor-mation (Security Rule compliance isrequired by April 21, 2005) or policychanges affecting how information isprocessed, organizations that deal with pro-tected health information in any form, willbe impacted by HIPAA.

In addition to protecting confidentialpatient information, the Privacy Rule was

By Christina Robbins, Marketing Program Manager, Digitech Systems

PaperVision Enterprise/ ImageSilo Feature

Disclosure Tracking Enhanced Auditing Force user to track critical information about every disclosure including a general purposedisclosure reason.

Disclosure Tracking Manual Disclosures Patient data released during a phone call or in office consultation can be tracked quickly and easily

Disclosure Tracking Document Grants Documents can be posted to a secure, web-accessible location combined with passwordsecurity for disclosure to out-of-office parties.

Protected Disclosure E-mail Auditing E-mails sent from the system are recorded in the system reports and the system can beconfigured to force a disclosure reason to be entered prior to sending the e-mail.

Protected Disclosure WorkFlow Can restrict security levels to allow individuals to see only the portions of a patient’s information that they need to perform their function.

Protected Disclosure Document-level Security Specific records can be filtered for viewing by specific individuals only.

De-identification Redaction User-level settings control whether or not a user can turn off redactions to reveal PHI.

De-identification Redaction Set redaction colors and styles so de-identified portions of a document blend seamlesslyinto the background.

De-identification WYSIWYG E-mail and Printing Redactions remain in place when e-mailing and printing

Patient Authorizations Imaging and Retrieval Scan and store every copy of a patient’s disclosure authorization form in a readily accessible format from any location, anytime, anywhere.

Security System Operations/Queries and Utilization Reports

Tracks each activity performed by any user accessing the system. See this informationdisplayed by user or by date/time.

Security System Operation/Queries Tracks not only activities successfully completed by the user, but all activities they tried toperform such as looking up a patient record. Allows the administrator to audit individualuser activity for anything that is suspicious.

Security User-level Security Set up user accounts to restrict the functions users have access to, such as email, printing, or annotations.

Security Document-level SecurityAudits

Perform security audits for individual documents. Provide a full listing of every access toand operation performed on the document.

Security Back-end Patient Data Secure, airtight document storage can easily be configured to allow access to the databaseand electronic document storage only through 128-bit encrypted caching technologies, so your database and document storage are never fully exposed to an end-user.

HIPAA Concern Advantage/Benefit

Digitech’s Solution for HIPAA-Enabling FunctionalityPaperVision Enterprise and ImageSilo are flexible document and content management solutions that will significantly assist any healthcare organization in attaining HIPAA

compliance. With their powerful feature sets, PaperVision and ImageSilo address the specific concerns brought to the forefront by the Privacy Rule changes in the healthcareindustry. The table matches key features and benefits provided by PaperVision Enterprise and ImageSilo with the specific HIPAA concerns they help to address.

In addressing the Security Rule, both systems provide tools that protect person-ally identifiable health information during electronic transmissions. These include e-mail and document grants which speed up the exchange of healthcare informationwhile ensuring the security and privacy of the document. For example, a doctor’soffice can e-mail an insurance company using a “document grant,” which would con-tain a URL web link to a specific document, with a pre-determined automatic expira-tion date and password. Since all online document distribution takes place oversecure 128 bit encrypted (SSL) technology, full HIPAA compliance is achieved whilealso enabling fast, secure and effective communication with the insurance company.

Feature-Rich Document ManagementIn addition to these HIPAA-specific features, PaperVision Enterprise and ImageSilo

are robust, scalable, feature-rich document and content management systems thatwill simplify document management needs in any organization while saving themmoney. In addition to traditional, scanned paper documents, the system can auto-matically import electronic files from an identified directory through the DirectoryManager component. These files are then available for retrieval using all of the fea-tures and functionality of the full document and content management system, andcan even be opened in their native file format for editing. Powerful search and retrievalcapabilities allow a user to access any document in the system, from any workstation,

at anytime. All types of standard business processes can be automated using theoptional workflow component. Processes are easy to set up using the graphical inter-face, and items will automatically appear on a user’s desktop. An ApplicationProgramming Interface (API) toolkit, included at no extra cost, allows the system toeasily integrate with other business-critical systems already in place. Additionally, thebasic system can be deployed in a matter of hours—not days or weeks like many otherelectronic document and content management systems. PaperVision Enterprise andImageSilo not only enable HIPAA compliance, but also save healthcare organizationstime and money. ❚

Digitech Systems, Inc. is a recognized leader in the development of software and services in the docu-ment and content management industry.We are committed to bringing the prospect of any document,any-where,anytime™ to an affordable reality.Digitech Systems, Inc.is headquartered in suburban Denver,CO.Thesoftware products provide affordable document and content management to organizations of any size.Threeprimary products include PaperFlow™ for image capture, PaperVision Enterprise for document and contentmanagement, and ImageSilo for secure off-site electronic document and content storage and retrieval.Theline can be purchased as a fully integrated suite, or as process-specific components. Ease-of-installation,extensive functionality, architectural flexibility, low price points and legendary toll free customer and technical support distinguish Digitech in the marketplace. For more information, or to obtaina free software demo CD, visit www.digitechsystems.com.

Supplement to November/December 2004 S9KMWorld

Supplement to November/December 2004S10 KMWorld

shop owners covering their windows withplywood and raising their inventory off the floor before a hurricane, signal the collision and interaction of primal marketforces.

What’s Going On?Why is this happening? Does it mean

anything? And should you care? The simpleanswers are: The value of content is chang-ing. With increasing volumes to consider, itisn’t just about content anymore. And, yes,because these changes are just the tip of theiceberg. Content-analysis technologies making their way into the marketplace willsignificantly reshape the processes of gathering, analyzing and communicatingthe intelligence that drives decisions. Thesenew procedures for analyzing unstructuredcontent allow companies to bring structureand standard reporting techniques to boththe constant flow and their deep archives ofunstructured information.

The forces driving changes in the valueof information are well known: volume, andthe continuing drive to deploy low-latency,highly responsive processes in every organi-zation. These forces have created a persistentdrumbeat for what some refer to as the “real-time organization.” Accelerating the decisionprocess in the face of overwhelming increasesin information volume creates a genuinedilemma—how to interpret more informationin less time.

This dilemma is complicated by a thirdforce—a perception of increasing uncertainty.Whether viewed from the perspective of theorganization, the market or industry, or globalconditions, the perceived risk of overlookingmeaningful information is clearly increasing,reflecting greater risks from geopolitics andincreasingly information-intensive industries.

What Does It Mean?In most practical applications, inter-

preting meaning is complicated by thevariability found in commonly held sensesof meaning. Variability is inherent in

complex concepts with multiple “dimen-sions” of meaning, e.g. brand value has at least two dimensions of meaning (customer satisfaction and willingness torecommend). Multiple perspectives alsocreate variability—from the differences inmeaning, or connotation, present in anygroup of individuals. For example, yourconcept of threat may reflect a strongersense of scale (massive vs. personal) thanmine—which may reflect greater sensitivityto immediacy.

Methods and procedures that do notaccount for, nor explicitly treat, the sourcesof variation will be less efficient (statisti-cally speaking, show greater variance) andmay well be biased (statistically, inevitably“off-target”). Methods that incorporatethese sources of variation into the interpre-tation process will also be capable ofaddressing important operational questions,such as “How strong is the consensus opin-ion of a team of analysts?” or “Among agroup of possible meanings, which is best supported by the evidence?” In short,ignoring these sources of variation increasesthe risks of simplification, lack of imaginationand substituting collective for collaborativethinking.

Recently, a large European servicesprovider employed this strategy to activate thepower of customer loyalty. Their first insightwas to distinguish two sources of loyalty vari-ation—customer satisfaction and willingnessto recommend. Next, using a “semantic ori-entation” technique recently summarized byTurney and Littman, they created a scoringengine to identify extremes of satisfaction andrecommendation. Integrated into an enter-prise search and alerting platform, they analyzed volumes of text feedback fromsources heretofore ignored. The results—bet-ter monitoring (“brand dashboards”) andtimely new dialog (“customer early warn-ing”)—helped convert potential brand energy(satisfaction) into kinetic brand energy (rec-ommendation).

Interpreting Meaning Adds ValueTremendous increases in the flow of

unstructured information have left manyorganizations and their intelligence processesill-prepared for both threat and opportunity.Without a means to rapidly assess and communicate meaning, the risk of missingsomething really important increases substantially. ❚

Fast Search & Transfer's FAST™ Marketrac™ is the first exampleof a new breed of search-derivative applications (SDA’s) that allowsusers to extract the meaning and trends from structured and unstruc-tured data. The underlying technology for interpreting meaning hasapplications far beyond market intelligence. FAST Marketrac is one ofthe first solutions to demonstrate tunable, repeatable techniques foradding structure to the mountains of information that are building up oncorporate servers all over the world.

Winds of Change

Extracting Meaning from Content

Over the last four years, the Internet haschanged many tried-and-true businesspropositions. From category-killers likeAmazon, to new-paradigms like eBay andPriceline, to new-media giants Google andYahoo!—this new medium continues to create new messages about business. Its central impact involves more than removing“the middle” of transactions; it has literallycreated new forums and new forms ofspeech—both commercial and non-com-mercial. With cyclonic force, it has becomethe most common publishing platform onthe planet. In its wake, it has left conven-tional thinking and traditional processesdevastated.

The changing stakes of an increasinglyWeb-centric world are clearly evidenced byrecent evolution in the market for businessinformation—the raw material of the gather-ing/analyzing/communicating paradigmcommonly referred to as “intelligence.”Historically, the market for information wasdriven by a combination of intrinsic content-worth and managed, exclusive access.Moreover, the economic model of publishersembodied this value proposition—informationis the critical fuel for decisions, better fuelproduces better decisions, therefore premiuminformation (better, more exclusive fuel) hasintrinsically greater value.

Publishers evolved new media—orforms of distribution gradually—carefullymanaging the balance between obsolescenceand innovation. These efforts, like those of

By Bjorn Olstad, CTO and Lee Phillips, Director, New Business Development

Fast Search & Transfer, Inc.™ (FAST™)

“The Internet has left

conventional thinking

and traditional

processes devastated.”

Supplement to November/December 2004 S11KMWorld

following symptoms: greater technical complexity, faster product cycles, increasedcustomer demands, limited pools of qualifiedand experienced employees, and ever-increasing productivity and profitabilityrequirements. The stress associated with considering these pressures is probablyenough on its own to warrant a good look atimplementing a KM initiative. KM can helprelieve you of the ever-increasing pressuresof today’s work environment.

3. KM benefits your budget (in an upor down economy).

In good economic times, companiesexperience rapid growth, and we all knowthat hiring, training and bringing newemployees up to speed are difficult chal-lenges. KM can accelerate a trip throughthe learning curve for all employees andenable them to expand their ability to performin areas beyond their traditional roles.

In poor economic times, companies areforced to closely watch expenses, operatewith skeleton staffs and maintain productivitymeasurements. A robust KM system canallow a company to handle greater workloads,increase productivity and improve profitabilitywithout increasing staff size. KM is the ultimate hedge fund of projects.

4. Delaying a KM implementation putsyou at a competitive disadvantage.

The major competitors in your industrymay already have a KM solution in place, orthey may be working on it now. As such,many of your competitors are already taking

steps toward becoming more effective,efficient and innovative. Their service levelsare increasing while their support costs areshrinking. How long can you afford to wait?

What’s more, many of the companiesthat have already recognized the benefits ofa KM initiative in one area are now lookingto expand the initiative to the rest of theenterprise. KM is giving your competitorsan advantage.

5. KM self-service saves money andmakes customers happy.

Certainly you’ve heard of self-service,and your company may already have it inplace in some form. Knowledge-enabledself-service is allowing many companiesto realize huge benefits. It makes a lot ofsense, because some problems or queriesare best suited for self-service and shouldnot be handled by a busy call centeragent.

For example, if I am a customer lookingfor a device driver for product XYZ, thenneither of us really wants to deal with theextra administration of a phone call. I wantto simply find the driver and download it.You want to allow me to obtain the driverthat way, because it is at least $10 cheaperto have me self-serve rather than call in andtalk with a live agent.

Now imagine a similar scenario inyour company where a knowledge-enabledself-service solution would be cheaperand more efficient than a live supportagent. Repeat this scenario hundreds ofthousands of times. KM will enable bothyour savings and your customers’ satisfactionto grow exponentially. So what are youwaiting for?

Make KM a priority in the call center,help desk and across the enterprise. Theextensive benefits derived from a successfulKM initiative can dwarf the incrementalimprovements that many other projects canbring. KM will help save you money, makecustomers happy, position you competitively,benefit your budget, address your growingknowledge needs and preserve your intel-lectual capital. KM will ultimately change,for the better, the way your enterprise performs. ❚

ServiceWare is a leading provider of knowledge-powered customer service and support solutions that improve agent productivityand customer satisfaction while reducing operating costs.ServiceWare’s enterprise and departmental solutions enable all businesses to easily manage,share and access corporate knowledge toeffectively answer inquiries through contact center, help desk or Webself-service interactions.

To learn more about ServiceWare solutions and services, visitwww.serviceware.com.

As the director of business analysis at ServiceWare, Tom Tobinhelps organizations evaluate the need for KM in their business centers.Tom has an MBA from the College of William & Mary, and a BS fromthe State University of New York. He is currently working on anadvanced degree in KM at George Washington University.

The Top 5 Reasons YouNeed KM—Today

There are undoubtedly many projects onyour corporate wish list, all tagged as “highpriority.” Many of the projects sound good,some seem interesting and they all promiseincredibly fast ROI. So how do you prioritizeyour opportunities and select the project thatwill deliver the most value?

While all of the projects you are con-sidering offer varying levels of incrementalreturns, the difference with a knowledgemanagement initiative is that it can offer

exponential returns, due to its very strategicnature. Here are some compelling reasonsto put it at the top of your list:

1. KM helps you capitalize on intellec-tual capital.

People (a.k.a. intellectual assets) areusually the first thing executives will citeas the true value of their company becauseof the detailed knowledge and hands-on experience workers possess in all areas of operations—technical, procedural,cultural, etc.

Intellectual capital is also first on our listbecause we agree that people are indeed the most treasured and critical assets of acompany. If this is true, then wouldn’t it beprudent to implement a system that couldharvest, store, protect, retrieve and distributethe knowledge that your people possessacross the enterprise—as soon as possible?

2. KM addresses your growing knowl-edge needs.

If you’re like most of us, your company isexperiencing at least some, if not all, of the

By Tom Tobin, Director of Business Analysis, ServiceWare Technologies

“KM will enable both your savings

and your customers’ satisfaction to

grow exponentially.

So what are you waiting for?”

Supplement to November/December 2004S12 KMWorld

quality and functionality of the system.Rarely do users think to measure themselvesagainst the quality of data they contribute.Projects are often considered nearly completewhen only a handful of tests have been con-ducted on a small subset of the knowledge.Moreover, the format of the results is beingjudged, not the ability of the results to solveproblems.

2. Tightly integrate with existing oper-ational processes. Nearly every knowledgemanagement system claims to report onknowledge effectiveness. In fact, most effec-tiveness data comes from user self-reporting,an unreliable method at best. Think about it:is a user likely to return to a knowledgerepository and confirm that a particular document helped them complete an RFP orsolve their user’s technical problem? Theonly way to measure knowledge effective-ness is by tightly coupling the knowledgebase with core operational systems.

3. Fully leverage knowledge in its “as-created” form. Only a fraction of anenterprise’s knowledge can be formatted inany standardized way to be useful, so it iscritical to leverage knowledge in whateverform it exists. The vast majority of technicalknow-how exists in case histories, technicaldocumentation, white papers, discussionforums, design documents and project deliverables. If even 1% of this knowledge isharvested into a structured form for knowledgemanagement, an organization ranks wellahead of its peers. The imperative to fullyleverage the remaining 99% is clear.

Technology or Business Challenge?To what extent does success in these

three areas depend on technology? Mostwould assume very little, but they would beunderestimating the potential for techno-logical improvement.

Workflow solutions are the traditionalapproach to bringing high-quality informationinto the knowledge base. They can incremen-tally improve productivity and support a larger number of contributors, performingmore tasks. But capturing even 10% of anextended enterprise’s knowledge into a struc-

tured knowledge base is still beyond reach.There is enormous potential for a technologyto automate the transformation of documentsinto structured knowledge.

Integrating with existing operational systems, while technologically demanding,creates a secondary, important issue—howthe user will interact with the system.Consider the merger of a knowledge man-agement and a call-tracking system whereinusers describe problems, are presented withsolutions and request assistance. The inter-face is not an out-of-the-box applicationscreen, but something heavily customized tointegrate with existing call-tracking systems.

Finally, the problem of leveraging knowl-edge in diverse formats cries out for morethan a search engine. Given a large set oflong, complex documents, search enginessimply indicate which ones are relevant.There are two profound limitations to thisapproach:◆ Relevance should be measured at a much

lower level of granularity than the docu-ment; and

◆ Results should be returned at a much lowerlevel of granularity than the document.A system which understood the fine-

grained structure of a document—its sections,chapters, paragraphs, etc.—could do dra-matically better.

Conclusion

It is clear that the technology underlyinga knowledge management solution isextremely important. First, the degree ofautomation with which disparately formattedinformation can be transformed into astructured knowledge base will be a criticaldeterminer of the quality and breadth ofinformation available. Second, the integrationrequired with transactional systems toachieve accurate feedback on knowledgequality renders most packaged user interfacesirrelevant; however, this means the projectwill be largely a custom development,opening an array of technology choices.Finally, even with strong transformationcapabilities, most knowledge will stillreside in a variety of original documentformats, in dire need of a better solutionthan today’s text search engines. ❚

Max Schireson is responsible for ensuring successful customerdeployment of Mark Logic products. This includes the delivery of professional services as well as the creation of value-added applicationsand solutions built on Mark Logic’s technology.

Mark Logic provides the first enterprise-class database for disparately structured content. Mark Logic Content Interaction Serverenables enterprises to analyze, synthesize and enhance business content locked inside Microsoft Word, PDF, e-mail, Excel, PowerPointand HTML, SGML and XML documents. Content Interaction Serverprovides the platform for a new class of content-centric applications.For more information, visit www.marklogic.com.

Does Technology Matter forKnowledge Management?

When knowledge management imple-mentations fail, the blame usually falls onchange management, business process orsome other soft factor. After all, the proxi-mate cause of the failure is most typically ashortage of good knowledge in the system.Why then would technology be to blame?

Knowledge management implementa-tions face a variety of environmental, oper-ational and cultural challenges:

Environmental—If the key challengeis capturing the knowledge of the extendedenterprise, then a diversity of formats andsources is intrinsic to the problem of knowl-edge management. Attempts to re-engineerthe way knowledge is generated in order to standardize how it is captured are gener-ally unsuccessful.

Operational—Knowledge is constantlyevolving as new solutions are identified,sometimes replacing the old solution oraugmenting it. The progress of understandingis irregular and complex. What begins as asingle problem can develop multiplecases—some with well-understood solutions,others with new solutions and still othersthat remain unsolved.

Cultural—The best knowledge is oftencreated in response to very real and concretechallenges, for example, mapping capabilitiesto customer needs. Contributing knowledgeto a corporate repository, while important toa knowledge manager, is simply not a highpriority when faced with the next urgent task.Answers created in an artificial context just to populate a knowledge base are, bycomparison, sterile and abstract.

Key Success FactorsGiven the breadth of challenges facing a

knowledge management implementation,teams clearly need to do some homework.Examining both successful and failed imple-mentations yields three factors common tothe successful projects:

1. Inject high-quality knowledge intothe knowledgebase. Striving for qualitymay seem obvious but it is often hard tomaintain this simple focus. IT delivers thesystem while users measure IT against the

By Max Schireson, Vice President, Customer Solutions, Mark Logic

Supplement to November/December 2004 S13KMWorld

Four facets of 10 nodes each have thesame discriminatory power as one taxono-my of 10,000 nodes.

That’s stunning. That means that withfacets, I can describe a collection with 40nodes (aka subject categories) that wouldtake a taxonomy 10,000 nodes to describe.That’s for an idealized case, of course, butthe gist of it holds true in the real world.The bottom line is that with facets, we canmake do with orders of magnitude fewercategories than we needed in a taxonomy

That’s because taxonomies are a type ofpre-coordinate indexing, meaning that itsbuilder anticipates the compound subjectspeople can browse along later, like “18thCentury French History.” In contrast, faceted

navigation is based on post-coordinateindexing, meaning that end-users stringtogether their own compound subjects atsearch time. They do this by combining sim-ple elements from multiple facets, in thisexample, (Time: 18th Century) + (Country:France) + (Topical Subject: History).

Reducing the number of categories weneed by orders of magnitude leads directlyto two primary benefits:

(1) Faceted navigation helps users moreeasily find what they’re looking for.

First, and most simply, it’s much easierfor users to grok 40 nodes than 10,000, soin practice, they start browsing more. Moreimportantly, faceted navigation offers usersmany different paths to each item of con-tent—often dozens or hundreds of paths.

For example, before Barnes &Noble.com deployed Endeca, users couldbrowse books through a taxonomy that

offered roughly 250,000 paths. Today, withtheir “BookBrowser,” Barnes & Noble offerstheir customers literally billions of paths tobrowse for books. That difference betweenless than a million paths, and billions ofpaths, is the incremental value facets bring.

(2) Faceted navigation lets contentowners streamline information manage-ment processes.

Most obviously, if you need orders ofmagnitude fewer categories or nodes thanbefore, it becomes exponentially simpler tomanage them. This leads to cascading bene-fits. It makes the work of auto-classifiers sim-pler, since they have fewer buckets to pickfrom. Also, facets operate independently ofeach other, which leads to a “schemaless”data model. This makes it simple to combineheterogeneous collections, because you don’tneed to mesh facets into a single taxonomy,and it makes it simple to add facets incre-mentally over time.

Guiding Light

So what is the difference betweenfaceted navigation and Guided Navigation?Guided Navigation is the name for Endeca’sversion of faceted navigation. It’s a meet-ing of classic information science, break-through computer science and gritty, real-world software tools and business process-es. Guided Navigation wraps search resultsin a context that shows users how to refineand explore their results, while constantlyremoving dead ends. With more than 200commercial deployments now, Endeca hascontinually built its expertise with leaderslike IBM, Barnes & Noble, and the Libraryof Congress back into the best practicesreflected in the tools, UIs and APIs that givecontent owners editorial flexibility andmanagerial control over their sites, turningfaceted navigation into a solution to a busi-ness problem.

Phew. That’s a lot of ground for a primer.To learn more about facets and GuidedNavigation, I recommend the first issue ofPhilip C. Murray’s Barrington Report onAdvanced Knowledge Organization andRetrieval. ❚

Endeca, headquartered in Cambridge, Massachusetts, wasfounded in 1999 to transform the online search and navigation experience so that people can easily access the full breadth and depthof large data sets. Today, Endeca solutions for enterprise search andcommerce are already helping businesses across a variety of sectorsincluding financial services, manufacturing, retail, informationproviders and business-to-business with applications that address the information overload problems associated with enterprise informationaccess and retrieval and content and catalog management.

Steve Papa is the CEO and co-founder of Endeca. Before startingEndeca, he was an early employee at the search engine companyInktomi, where he worked with unstructured data on a massive scale,and before that, at the data warehouse company Teradata, where heworked on structured data on a massive scale.

A Primer on FacetedNavigation and GuidedNavigation

“The future is faceted.”—Peter Morville,president and founder, Semantic Studios

You’ve likely heard the buzz by now: facetednavigation truly changes the ground rules forKM, and for the many related applicationswhere users need to find information, rangingfrom subscription content services to directo-ries. It complements search and relevance rank-ing to fill a big hole in the process of makingknowledge reusable. Faceted navigationbrings the double whammy of helping usersfar more easily find what they’re looking for,while also helping content owners to far more efficiently manage content (or in reality, to make it even feasible to managecontent at all.)

Below, I’ll talk about the distinct proper-ties of faceted navigation that create thosebenefits, and later, explain how classicinformation science combined with break-throughs in computer science bring it to lifein a form called “Guided Navigation.”

What is faceted navigation? It’s a way tobrowse information, or to refine long lists ofsearch results, along multiple dimensions,aka facets. These are orthogonal lensesthrough which to view the world. For exam-ple, I might search for an expert by facetslike name, project, company, or date—andmore likely, by some combination of thosefacets, selected in any sequence.

So how do facets work their powers?First, we need to state what I’ll call Busch’sgolden law of facets, named for JosephBusch of Taxonomy Strategies, a past president of the American Society forInformation Science:

By Steve Papa, CEO and Co-Founder, Endeca

“With facets, we can make do with orders

of magnitude fewer categories.”

Supplement to November/December 2004S14 KMWorld

being able to answer almost any CRM-related question the moment it is asked.

Now, six months after your go-live date,end users and executives are still strugglingto answer daily business questions as youcontinue to search for a solution to betterempower them.

Sound familiar? You’re not alone. Mostcompanies find that empowering users is oneof the most difficult hurdles to overcome for asuccessful implementation. Accessing mean-ingful, relevant, critical data is far more diffi-

cult than anyone expected. The reality is thatthese powerful and complex enterprise appli-cations, intended to help you better manageyour daily business, require reporting tools tomaximize your return on investment.

Operational Reporting in the Real World

Leading industry analysts and scores ofsatisfied customers consistently rate Siebelat the top of the CRM market. However,Siebel’s data model makes it difficult fornon-technical users to identify all of theinformation stored in the application.Often, customers have data that is inacces-sible to business users. As a result, ITdepartments are forced to act as informationgatekeepers, incurring enormous expense tosolve the problem. Siebel has attempted tobridge this gap in their CRM applicationsby including a set of pre-built reports thatprovide a fixed view into the data. Butfixed-view reports often lack the detailneeded to fully understand a business issue,making it difficult to form decisions with-out all the relevant information.

Noetix addresses this problem by pro-viding customizable reporting for Siebel,automatically generating metadata fromenterprise applications and enabling imme-diate access to data. Generating thousandsof business views and pre-built reports thatmatch your unique Siebel configuration(including security, metadata and businessrules), the Noetix approach shields reportsfrom application upgrades and reconfigura-tions, providing users with uninterruptedoperational reporting and reducing theexpense of report maintenance.

Without such help, reporting fromSiebel requires developers with an in-depthunderstanding of the underlying tablestructures to access data, using difficult-to-use tools that have been designed for tech-nical users. Siebel business users can nowaccess the data they need in the same con-text and terminology as their Siebel screensright out-of-the-box, thus eliminating theheavy reliance upon IT resources. ❚

More than 1,200 customers worldwide use Noetix to quickly and costeffectively access the enterprise application data they need to makeimportant business decisions every day. Unlike tools that requireweeks of manual mapping to produce a single report, the Noetix solu-tion automatically generates metadata objects specific to an organi-zation’s unique configuration of Oracle or Siebel. Using a proprietaryauto-discovery process to “crawl” the application database, Noetixtransforms data into business content that is searchable and naviga-ble. Noetix dynamically generates meaningful and immediatelyusable business information that reflects an organization’s specificapplication configuration. Noetix customers include Motorola,Starbucks,Toshiba,VISA and Xerox International Partners.The compa-ny is headquartered in Bellevue, WA, with international operationsbased in London. For more information about Noetix, visit their Website at http://www.noetix.com or call toll free at 866-4NOETIX.

Maximize Your Investment

Empowering CRM Users with Knowledge

In accordance with your overall KM strategy,your organization has decided to invest in a customer relationship management (CRM) sys-tem. After a thorough vendor analysis, Siebelis chosen as the best solution to meet your or-ganization’s demanding business requirements.

One main objective is to provide opera-tional reporting for users and executivesthat doesn’t drain IT resources. Instead ofspending days compiling informationmanually (or worse, never being able toanswer the question at all), with the imple-mentation of Siebel, you imagined users

A Noetix Case Study

NoetixViews®—The Driving Force Behind NoetixNoetixViews unlocks the operational data within the Siebel CRM applications

by automatically translating Siebel user screens into easy-to-use virtual tablesusing the field names familiar to end users.

Noetix AnswerPoint—Immediate Answers to User QuestionsNoetix AnswerPoint provides end users with immediate answers to the most

common business questions. This component delivers a comprehensive library ofcustomizable report templates for Siebel CRM, automatically customized to eachapplication’s unique configuration.

Noetix QueryServer®—Centralized Data AccessNoetix QueryServer centralizes queries and user management, providing sim-

plified integration between Siebel CRM data and other enterprise data. By inte-grating disparate data sources into virtual tables, Noetix QueryServer gives queryusers a simple, single point of access to all enterprise data sources, regardless ofwhere the data resides.

Noetix WebQuery®—Simplified Web-Enabled AccessNoetix WebQuery, a simplified query tool, empowers end users to create and

modify reports without IT involvement. Noetix WebQuery easily integrates with enter-prise portals and works with popular reporting tools, including Microsoft Excel.

Noetix Benefits:◆ Delivers immediate answers to sales, customer service and marketing business

questions;◆ Empowers business professionals to make informed, data-driven decisions

using an ad hoc reporting solution;◆ Provides a flexible, Web-based reporting platform for faster, easier Siebel

reporting; and◆ Allows users to export queries into Microsoft Excel for detailed drilldown and

analysis.

The Noetix Solution for Siebel

Supplement to November/December 2004 S15KMWorld

tions, no off-the-point content ... just a high-impact source for useful information.

As an amplifier, KMWorld White Paperscan add volume to the message of a smaller,specialized vendor that you may not haveheard from yet. Many of the sponsors ofKMWorld White Papers don’t have the cloutor the resources to run massive PR cam-paigns and marketing programs. What theyDO have, however, are solutions that can bedeployed in your organization that will bring

competitive advantage and process improve-ment ... if only you could hear them.

And as a communications link, therejust isn’t any other publication that bringstogether buyers and sellers of informationmanagement tools such as these.

How It WorksThe KMWorld Specialty Publishing

group has created an editorial calendar (seebelow) for a full calendar year. Sponsorsare able to pinpoint specific issues wheretheir message will be in the proper contextto reach an audience in search of their kindof solution, and their kind of solution only.

Our editors review all submissions foradherence to strict editorial guidelines. Wedo not allow “hype.” We DO allow reasonedexplanations of their solutions, and how theycan help. As a “business solutions” paper, wehave every desire for our sponsors’ messagesto come through loud and clear.

Our editorial calendar is not immovable.If you have a suggestion for a new topic, ver-tical market or technology space to explore,please contact us. ❚

Andy Moore, andy_moore@verizon.netKathryn Rogals, kathy_rogals@kmworld.comPaul Rosenlund, paul_rosenlund@kmworld.com

KMWorld Best PracticesWhite PapersThe Whys and Wherefores of a Business Solutions Series

The KMWorld Best Practice White Paper Se-ries serves sometimes as a filter, sometimesas an amplifier and sometimes as a simplecommunications link between those whohave business goals to reach, and those whohave the solutions to help achieve those goals.

As a filter, KMWorld White Papers nar-row down to specific subject areas with eachissue—see below. You can focus on one sub-ject at a time, and immerse yourself in thebest solutions available to you from the larg-er “technology” marketplace. No distrac-

The 2005 “Best Practices” Series...JanuaryBusiness Process Management

FebruaryFinancial Services

MarchE-mail & IM Management

AprilEnterprise Search

MayEnterprise Content Management

JuneGovernment

July/AugustImaging & Forms Processing

SeptemberRecords Management

OctoberCollaborative Knowledge Sharing

November/DecemberKnowledge Management

When You’re a Global Business Driver,It’s Good to Have an Owner’s Manual

KMWorld Best Practices White Papers ♦ Powerful ♦ Direct ♦ Objective ♦ Content-rich ♦

By the KMWorld Specialty Publishing Group

www.infotoday.com

Produced by:

KMWorld MagazineSpecialty Publishing Group

For information on participating in the next white paper in the “Best Practices” series. contact:paul_rosenlund@kmworld.com or kathy_rogals@kmworld.com • 207.338.9870

Kathryn Rogals Paul Rosenlund Andy Moore207-338-9870 207-338-9870 207-236-0331kathy_rogals@kmworld.com paul_rosenlund@kmworld.com andy_moore@verizon.net

For more information on the companies who contributed to this white paper, visit their Web sites or contact them directly:

Digitech Systems Inc.8400 East Crescent ParkwaySuite 600Greenwood Village CO 80111

PH: 888.374.3569FAX: 866.245.3569E-mail: info@digitechsystems.comWeb: www.digitechsystems.com

Endeca Technologies, Inc.55 Cambridge ParkwayCambridge MA 02142

PH: 617.577.7999FAX: 617.577.7766E-mail: sales@endeca.comWeb: www.endeca.com

Fast Search & Transfer, Inc.

117 Kendrick Street, Suite 100Needham MA 02494

PH: 888.871.3839FAX: 781.304.2410E-mail: fastresponse@fastsearch.comWeb: www.fastsearch.com

Mark Logic Corporation2000 Alameda de las PulgasSuite 100San Mateo CA 94403

PH: 650.655.2300FAX: 650.655.2310 E-mail: sales@marklogic.comWeb: www.marklogic.com

Hummingbird Ltd.1 Sparks AvenueToronto ON M2H 2W1

PH: 877.FLY.HUMM or416.496.2200

FAX: 416.496.2207E-mail: getinfo@hummingbird.comWeb: www.hummingbird.com

Convera1921 Gallows Road, Suite 200Vienna VA 22182

PH: 800.788.7758 or 703.761.3700

FAX: 703.761.1990E-mail: info@convera.comWeb: www.convera.com

ServiceWare Technologies, Inc.12 Federal Street, Ste 503 Pittsburgh PA 15212

PH: 412.222.4450 or 800.572.5748

FAX: 412.222.4455E-mail: info@serviceware.comWeb: www.serviceware.com

Noetix Corp229 112th Avenue NE Bellevue WA 98004

PH: 866.4NOETIXFAX: 425.372.2942E-mail: sales@noetix.comWeb: www.noetix.com

www.kmworld.com