XML Technologies & Content Strategies

If you have been following recent XML Technologies blog entries, you will notice we have been talking a lot lately about XML Smart Content, what it is and the benefits it can bring to an organization. These include flexible, dynamic assembly for delivery to different audiences, search optimization to improve customer experience, and improvements for distributed collaboration. Great targets to aim for, but you may ask are we ready to pursue these opportunities? It might help to better understand the technology landscape involved in creating and delivering smart content.

The figure below illustrates the technology landscape for smart content. At the center are fundamental XML technologies for creating modular content, managing it as discrete chunks (with or without a formal content management system), and publishing it in an organized fashion. These are the basic technologies for "one source, one output" applications, sometimes referred to as Singe Source Publishing (SSP) systems.

The innermost ring contains capabilities that are needed even when using a dedicated word processor or layout tool, including editing, rendering, and some limited content storage capabilities. In the middle ring are the technologies that enable single-sourcing content components for reuse in multiple outputs. They include a more robust content management environment, often with workflow management tools, as well as multi-channel formatting and delivery capabilities and structured editing tools. The outermost ring includes the technologies for smart content applications, which are described below in more detail.

It is good to note that smart content solutions rely on structured editing, component management, and multi-channel delivery as foundational capabilities, augmented with content enrichment, topic component assembly, and social publishing capabilities across a distributed network. Descriptions of the additional capabilities needed for smart content applications follow.

Content Enrichment / Metadata Management: Once a descriptive metadata taxonomy is created or adopted, its use for content enrichment will depend on tools for analyzing and/or applying the metadata. These can be manual dialogs, automated scripts and crawlers, or a combination of approaches. Automated scripts can be created to interrogate the content to determine what it is about and to extract key information for use as metadata. Automated tools are efficient and scalable, but generally do not apply metadata with the same accuracy as manual processes. Manual processes, while ensuring better enrichment, are labor intensive and not scalable for large volumes of content. A combination of manual and automated processes and tools is the most likely approach in a smart content environment. Taxonomies may be extensible over time and can require administrative tools for editorial control and term management.

Component Discovery / Assembly: Once data has been enriched, tools for searching and selecting content based on the enrichment criteria will enable more precise discovery and access. Search mechanisms can use metadata to improve search results compared to full text searching. Information architects and organizers of content can use smart searching to discover what content exists, and what still needs to be developed to proactively manage and curate the content. These same discovery and searching capabilities can be used to automatically create delivery maps and dynamically assemble content organized using them.

Distributed Collaboration / Social Publishing: Componentized information lends itself to a more granular update and maintenance process, enabling several users to simultaneously access topics that may appear in a single deliverable form to reduce schedules. Subject matter experts, both remote and local, may be included in review and content creation processes at key steps. Users of the information may want to "self-organize" the content of greatest interest to them, and even augment or comment upon specific topics. A distributed social publishing capability will enable a broader range of contributors to participate in the creation, review and updating of content in new ways.

Federated Content Management / Access: Smart content solutions can integrate content without duplicating it in multiple places, rather accessing it across the network in the original storage repository. This federated content approach requires the repositories to have integration capabilities to access content stored in other systems, platforms, and environments. A federated system architecture will rely on interoperability standards (such as CMIS), system agnostic expressions of data models (such as XML Schemas), and a robust network infrastructure (such as the Internet).

These capabilities address a broader range of business activity and, therefore, fulfill more business requirements than single-source content solutions. Assessing your ability to implement these capabilities is essential in evaluating your organizations readiness for a smart content solution.

Over the past few weeks, since publishing Smart Content in the Enterprise, I’ve had several fascinating lunchtime conversations with colleagues concerned about content technologies. Our exchanges wind up with a familiar refrain that goes something like this. “Geoffrey, you have great insights about smart content but what am I supposed to do with all this information?” Ah, it’s the damning with faint praise gambit that often signals an analysis paralysis conundrum for decision-making.

Let me make one thing perfectly clear -- I do not have an out-of-the-box prescription for a solution. It’s not simply a matter of focusing on your customer experience, optimizing your content for search, investing in a component content management platform, or adopting DITA – although, depending on the situation, I may recommend some combination of these items as part of a smart content strategy.

For me, smart content remains a work in progress. I expect to develop the prescriptive road map in the months ahead. Here’s a quick take on where I am right now.

• For publishers, it’s all about transforming the publishing paradigm through content enrichment – defining the appropriate level of granularity and then adding the semantic metadata for automated processing.
• For application developers, it’s all about getting the information architecture right and ensuring that it’s extensible. There needs to be sensible storage, the right editing and management tools, multiple methods for organizing content, as well as a flexible rendering and production environment.
• For business leaders and decision makers, there needs to be an upfront investment in the right set of content technologies that will increase profits, reduce operating costs, and mitigate risks. No, I am not talking about rocket science. But you do need a technology strategy and a business plan.

As highlighted by the case studies included in the report, I can point to multiple examples where organizations have done the right things to produce notable results. Dale and I will continue the smart content discussions at the Gilbane Boston conference right after Thanksgiving, both through our preconference workshop, and at a conference session “Smart Content in the Real World: Case Studies and Real Results.”

We are also launching a Smart Content Readiness Service, where we will engage with organizations on a consulting basis to identify:

• The business drivers where smart content will ensure competitive advantage when distributing business information to customers and stakeholders
• The technologies, tools, and skills required to componentized content, and target distribution to various audiences using multiple devices
• The operational roles and governance needed to support smart content development and deployment across an organization
• The implementation planning strategies and challenges to upgrade content and creation and delivery environments

Please contact me if you are interested in learning more.

In short, to answer my lunchtime colleagues, I cannot (yet) prescribe a fully baked solution. It’s too early for the recipes and the cookbook. But I do believe that the business opportunities and benefits are readily at hand. At this point, I would invite you to join the discussion by letting me know what you expect, what approaches you’ve tried, where you’ve wound up, what you think needs to come next – and how we might help you.

What is hot in XML these days? I have been to a few conferences and meetings, talked with many clients, participated in various research projects, and developed case studies on emerging approaches to XML adoption. DITA (Darwin Information Typing Architecture) is hot. Semantically enriched XML is hot. Both enable some interesting functionality for content delivered via print, on the web, and through mobile delivery channels. These include dynamic assembly of content organized into a variety of forms for custom uses, improved search and discovery of content, content interoperability across platforms, and distributed collaboration in creating and managing content.

On November 30, prior to the Gilbane Conference in Boston, Geoff Bock and I will be holding our 3rd workshop on Smart Content which is how we refer to semantically enriched, modular content (it's easier to say). In the seminar we will discuss what makes content smart, how it is being developed and deployed in several organizations, and dive into some technical details on DITA and semantic enrichment.  This highly interactive seminar has been well received in prior sessions, and will be updated with our recently completed research findings.  More information on the seminar is available at  http://gilbaneboston.com/workshops.html.

By the way, the research report, entitled Smart Content in the Enterprise, is now available at the research section at Gilbane.com. It includes several interesting case studies from a variety of organizations, and a lot of good information for those considering taking their content to the next level. We encourage you to download it (it is free). I also hope to see you in Boston at the workshop.

Traditionally, publishing is a pushy process. When I have something to say, I write it down. Perhaps I revise it, check with colleagues, and verify my facts with appropriate authorities. Then I publish it, and move on to the next thing – without directly interacting with my audience and stakeholders. Whether I distribute the content electronically or in a hard copy format, I leave it to my readers to determine the value of whatever I publish.

However, as we describe in our recently completed report Smart Content in the Enterprise, XML applications can transform this conventional publishing paradigm. By smart content, we mean content that is granular at the appropriate level, semantically rich, useful across applications, and meaningful for collaborative interaction.

From a business perspective, smart content adds value to published information in new and compelling ways. Let’s consider the experiences of NetApp and Warrior Gateway, two of the organizations featured in our report.

NetApp
As a provider of storage and data management solutions, NetApp has invested a lot of time and effort embracing DITA and restructuring its technical documentation. By systematically tagging and managing content components, and by focusing on the underlying content development processes, writers and editors can keep up with the pace of product releases.

But there is more to this publishing process orientation. Beyond simply producing product information faster and cheaper, NetApp is poised to make publishing better. The company can now easily support its reseller partners by providing them with the DITA tagged content that they can directly incorporate into their own OEM solutions. Resellers' customers get just the information they need, directly from the source. With its XML application, NetApp incorporates its partners and stakeholders into its information value chain.

Warrior Gateway
As a content aggregator, Warrior Gateway collects, organizes, enriches, and redistributes content about a wide range of health, welfare, and veteran-related services to soldiers, veterans, and their families. Rather than simply compiling an online catalog of service providers’ listings, Warrior Gateway restructures the content that government, military, and local organizations produce, and enriches it by adding veteran-related categories and other information. Furthermore, Warrior Gateway adds a social dimension by encouraging contributions from veterans and family members.

Once stored within the XML application powering Warrior Gateway, the content is easily reorganized and reclassified to provide the veterans’ perspective about areas of interest and importance. Volunteers working with Warrior Gateway can add new categories when necessary. Service providers can claim their profile and improve their own data details. Even the public users can contribute to content to the gateway, a crowd sourcing strategy to efficiently collect feedback from users. With contributions from multiple stakeholders, the published listings can be enriched over time without requiring a large internal staff to add the extra information.

Capturing New Business Value
There’s a lot more detail about how the XML applications work in our case studies – I recommend that you check them out.

What I find intriguing is the range of promising and potentially profitable business models engendered by smart content.  Enterprise publishers have new options and can go beyond simply pushing content through a publishing process. Now they can build on their investments, and capture the pull of content value.

Authoring in a structured text environment has traditionally been done with dedicated structured editors. These tools enable validation and user assisted markup features that help the user create complete and valid content. But these structured editors are somewhat complicated and unusual and require training in their use for the user to become proficient. The learning curve is not very steep but it does exist.

Many organizations have come to see documentation departments as a process bottleneck and try to engage others throughout the enterprise in the content creation and review processes. Engineers and developers can contribute to documentation and have a unique technical perspective. Installation and support personnel are on the front lines and have unique insight into how the product and related documentation is used. Telephone operators not only need the information at their fingertips, but can also augment it with comments and ides that occur while supporting users. Third-party partners and reviewers may also have a unique perspective and role to play in a distributed, collaborative content creation, management, review, and delivery ecosystem.

Our recently completed research on XML Smart Content in the Enterprise indicates that as we strive to move content creation and management out of the documentation department silo, we will also need to consider how the data is encoded and the usefulness of the data model in meeting our expanded business requirements. Smart content is multipurpose content designed with several uses in mind. Smart content is modular to support being assembled in a variety of forms. And smart content is structured content that has been enriched with semantic information to better identify it's topic and role to aide processing and searching. For these reasons, smart content also improves distributed collaboration. Let me elaborate.

One of the challenges for distributed collaboration is the infrequency of user participation and therefore, unfamiliarity with structured editing tools. It makes sense to simplify the editing process and tools for infrequent users. They can't always take a refresher course in the editor and it's features. They may be working remotely, even on a customer site installing equipment or software. These infrequent users need structured editing tools that are designed for them. These collaboration tools need to be intuitive and easy to figure out, easily accessible from just about anywhere, and should be affordable and have flexible licensing to allow a larger number of users to participate in the management of the content. This usually means one of two things: either the editor will be a plug in to another popular word processing system (e.g., MS Word), or it will be accessed though a thin-client browser, like a Wiki editor. In some environments, it is possible that both may be need in addition to traditional structured editing tools. Smart content modularity and enrichment allows flexibility in editing tools and process design. This allows the  use of a variety of editing tools and flexibility in process design, and therefore expanding who can collaborate from throughout the enterprise.

Also, infrequent contributors may not be able to master navigating and operating within a  complex repository and workflow environment either for the same familiarity reasons. Serving up information to a remote collaborator might be enhanced with keywords and other metadata that is designed to optimize searching and access to the content. Even a little metadata can provide a lot of simplicity to an infrequent user. Product codes, version information, and a couple of dates would allow a user to hone in on the likely content topics and select content to edit from a well targeted list of search results. Relationships between content modules that are indicated in metadata can alert a user that when one object is updated, other related objects may need to be reviewed for potential update as well.

It is becoming increasingly clear that there is no one model for XML or smart content creation and editing. Just as a carpenter may have several saws, each designed for a particular type of cut, a robust smart content structured content environment may have more than one editor in use. It behooves us to design our systems and tools to meet the desired business processes and user functionality, rather than limit our processes to the features of one tool.

One of the conclusions of our report Smart Content in the Enterprise (forthcoming next week) is how a little bit of enrichment goes a long way. It’s important to build on your XML infrastructure, enrich your content a little bit (to the extent that your business environment is able to support), and expect to iterate over time.

Consider what happened at Citrix, reported in our case study Optimizing the Customer Experience at Citrix: Restructuring Documentation and Training for Web Delivery. The company had adopted DITA for structured publishing several years ago. Yet just repurposing the content in product manuals for print and electronic distribution, and publishing the same information as HTML and PDF documents, did not change the customer experience.

A few years ago, Citrix information specialists had a key insight: customers expected to find support information by googling the web. To be sure, there was a lot of content about various Citrix products out in cyberspace, but very little of it came directly from Citrix. Consequently the most popular solutions available via web-wide searching were not always reliable, and the detailed information from Citrix (buried in their own manuals) was rarely found.

What did Citrix do? Despite limited resources, the documentation group began to add search metadata to the product manuals. With DITA, there was already a predefined structure for topics, used to define sections, chapters, and manuals. Authors and editors could simply include additional tagged metadata that identified and classified the contents – and thus expose the information to Google and other web-wide search engines.

Nor was there a lot of time or many resources for up-front design and detailed analysis. To paraphrase a perceptive information architect we interviewed, “Getting started was a lot like throwing the stuff against a wall to see what sticks.” At first tags simply summarized existing chapter and section headings. Significantly, this was a good enough place to start.

Specifically, once Citrix was able to join the online conversation with its customers, it was also able to begin tracking popular search terms. Then over time and with successive product releases, the documentation group was able to add additional tagged metadata and provide ever more focused (and granular) content components.

What does this mean for developing smart content and leveraging the benefits of XML tagging? Certainly the more precise your content enrichment, the more findable your information is going to be. When considering the business benefits of search engine optimization, the quality of your tagging can always improve over time. But as a simple value proposition, getting started is the critical first step.

In our recently completed research on Smart Content in the Enterprise we explored how organizations are taking advantage of benefits from XML throughout the enterprise and not just in the documentation department. Our findings include several key issues that leading edge XML implementers are addressing including new delivery requirements, new ways of creating and managing content, and the use of standards to create rich, interoperable content. In our case studies we examined how some are breaking out of the documentation department silo and enabling others inside or even outside the organization to contribute and collaborate on content. Some are even using crowd sourcing and social publishing to allow consumers of the information to annotate it and participate in its development. We found that expectations for content creation and management have changed significantly and we need to think about how we organize and manage our data to support these new requirements. One key finding of the research is that organizations are taking a different approach to repurposing their content, a more proactive approach that might better be called "multipurposing".

In the XML world we have been talking about repurposing content for decades. Repurposing content usually means content that is created for one type of use is reorganized, converted, transformed, etc. for another use. Many organizations have successfully deployed XML systems that optimize delivery in multiple formats using what is often referred to as a Single Source Publishing (SSP) process where a single source of content is created and transformed into all desired deliverable formats (e.g., HTML, PDF, etc.).

Traditional delivery of content in the form of documents, whether in HTML or PDF, can be very limiting to users who want to search across multiple documents, reorganize document content into a form that is useful to the particular task at hand, or share portions with collaborators. As the functionality on Web sites and mobile devices becomes more sophisticated, new ways of delivering content are needed to take advantage of these capabilities. Dynamic assembly of content into custom views can be optimized with delivery of content components instead of whole documents. Powerful search features can be enhanced with metadata and other forms of content enrichment.

SSP and repurposing content traditionally focuses on the content creation, authoring, management and workflow steps up to delivery. In order for organizations to keep up with the potential of delivery systems and the emerging expectations of users, it behooves us to take a broader view of requirements for content systems and the underlying data model. Developers need to expand the scope of activities they evaluate and plan for when designing the system and the underlying data model. They should consider what metadata might improve faceted searching or dynamic assembly. In doing so they can identify the multiple purposes the content is destined for throughout the ecosystem in which it is created, managed and consumed.

Multipurpose content is designed with additional functionality in mind including faceted search, distributed collaboration and annotation, localization and translation, indexing, and even provisioning and other supply chain transactions. In short, multipurposing content focuses on the bigger picture to meet a broader set of business drivers throughout the enterprise, and even beyond to the needs of the information consumers.

It is easy to get carried away with data modeling and an overly complex data model usually requires more development, maintenance, and training than would otherwise be required to meet a set of business needs. You definitely want to avoid using specific processing terminology when naming elements (e.g., specific formatting, element names that describe processing actions instead of defining the role of the content). You can still create data models that address the broader range of activities without using specific commands or actions. Knowing a chunk of text is a "definition" instead of an "error message" is useful and far more easy to reinterpret for other uses than an "h2" element name or an attribute for display='yes'. Breaking chapters into individual topics eases custom, dynamic assembly. Adding keywords and other enrichment can improve search results and the active management of the content. In short, multipurpose data models can and should be comprehensive and remain device agnostic to meet enterprise requirements for the content.

The difference between repurposing content and multipurpose content is a matter of degree and scope, and requires generic, agnostic components and element names. But most of all, multipurposing requires understanding the requirements of all processes in the desired enterprise environment up front when designing a system to make sure the model is sufficient to deliver designed outcomes and capabilities. Otherwise repurposing content will continue to be done as an afterthought process and possibly limit the usefulness of the content for some applications.
 

If you've been reading our recent posts on Gilbane's new research on XML adoption, you might be wondering how to get the report in advance of its availability from Gilbane later this month.

Smart Content in the Enterprise: How Next Generation XML Applications Deliver New Value to Multiple Stakeholders is currently offered by several of the study sponsors: IBM, JustSystems, MarkLogic, MindTouch, Ovitas, Quark, and SDL.

We'll also be discussing our research in real time during a webinar hosted by SDL on November 4. Look for details within the next few weeks.

If you're only reading this XML blog, be sure to check out my recent blog post Focusing on Smart Content, which I published in the main Gilbane blog.

We've published a new paper on addressing large-scale integration, storage, and access of complex information. As Dale mentions in his entry over on our main blog, the paper frames the discussion in terms of challenges to Open Government initiatives. We note, though, that the exploration of obstacles to effective, efficient processing of high volumes of data and content is relevant across many industries.

We're cross-posting here on the XML blog because the paper deals wtih XML content and the XML family of standards, including XQuery and XPath.

The Gilbane Beacon is available as a free download from Gilbane and from Mark Logic, sponsor of the paper.