Saturday, October 20, 2007

S1000D and SCORM Integration

I gave a presentation yesterday on S1000D and SCORM integration at the Doctrain 2007 conference in Lowell, MA. The main goal of integrating these two specifications is to reduce product life cycle costs and eliminate redundancies by streamlining business processes across the documentation and training functions. I noted that there are many opportunities for data reuse at every phase of the product life cycle including: concept, design, manufacturing, assembly, testing, delivery, and support. Documentation and training belong to the support phase and are often the only departments where content is captured in XML. This is due to the complexity and cost of current specialist XML authoring tools. XForms will allow knowledge workers to contribute knowledge assets in XML at every phase of the product lifecycle with a simple web form. I also believe that Office Open XML (OOXML) offers the opportunity to extract some value out of MS Office documents by exposing their contents to XML processing languages and tools such as XSLT 2.0 and XQuery.

Engineering data should be the trusted source of data for both publications and training. For example, product model data and engineering drawings can be used to create simulation for training and manufacturing assembly instructions can be used to create installation procedures for publications. Any data reuse strategy should look beyond training and publication to identify ways to reuse data and streamline processes across the entire product lifecycle.

The integration of SCORM and S1000D presents management and technical challenges. Since training and documentation are often two separate functions within the enterprise, the integration can have an impact on budget, processes, roles, and the organizational structure. For example, the cross-functional integrated project team is a good approach. The success of the integration will also require top leadership commitment and support. It is also important to address technical challenges such as the integration of existing content management systems (CMS) and learning management systems (LMS).

The first technical approach is to create dual purpose data modules (DMs). I highlighted the importance of clearly defining and documenting business rules, particularly for dual purpose data modules. The business rules should specify among other things the appropriate level of granularity and language style (e.g. Simplified English). These business rules should be validated with technologies such as ISO Schematron and Simplified English Checkers. Since S1000D is weak on learning content metadata, the IEEE Learning Object Metadata (LOM) specification should be used to add learning object metadata to the S1000D dual purpose DMs. It is also possible to package all training data modules as a training publication module (PM). XSLT is then used to transform the S1000D DM into SCORM sharable content objects (SCOs) and learning assets. The S1000D metadata (IDSTATUS) shall be retained in the result SCOs to facilitate product applicability filtering when the SCOs are presented to the learners with an LMS. The S1000D PM can be used to generate the SCORM manifest as well. The dual purpose S1000D DM approach does not always support complex learning interactions and good instructional design principles.

The second approach is to give complete freedom to the instructional designer to design an effective learning experience. All elements in the S1000D data modules that are reusable in SCOs are assigned a unique ID. Examples are: paragraphs, steps, warning, cautions, notes, tables, etc. This can be done automatically using the XSLT generate-id() function. The instructional designer then searches the CSDB to find and display relevant DMs. She can then use XInclude/XPointer to include reusable elements from the DM into the SCO. When this is done, the SCO is automatically updated when the DM is updated.

Future versions of the S1000D specification will incorporate change proposal forms (CPFs) that will facilitate the integration of SCORM and S1000D content.

A copy of my presentation is available here.

No comments: