Australian Taxation Office (ATO)

Improving the quality and efficiency of contributions from a large user base of Microsoft Word users.

Behind a large, dynamic, public website representing the Australian Tax Office (ATO) are a significant number of content contributors and maintainers. This page provides an overview of their processes.

Challenge

The Australian Tax Office (ATO) has tens of thousands of knowledge workers using a standardized desktop computing environment. While domain experts need to continually update documents on the ATO website, it is impractical to allow all staff to access the web content management system. To ensure website quality and integrity, all contributions relay through a publications team who post updates to the site.

This workflow created a problem familiar to anyone with production experience.
After contributors submit their content to the publications team, they often tend to keep writing. So by the time the content is ready to release, there is a new version available with none of the fixes from the publications team.

The problems with this workflow include:

Time invested by publications staff is often wasted,
The quality of contributions can vary widely which requires additional training to process,
The lead times required to process an update are longer,
Without feedback, contributors are unable to control the lifecycle of their own documents.

No one chooses an inefficient workflow out of preference, especially an experienced organization responsible for publishing a significant amount of content to a large population of consumers. With so many users and so much content, inefficiency adds up to significant costs.

Solution

To address this problem, Allette Systems created a service to allow end users to validate their Word files (docx format) for compliance with the ATO document template. This solution has made a profound difference to the efficiency of their publishing workflow.

Word does not provide users with the information necessary to understand where they have violated the rules of document templates. The conventional solutions have been a combination of locking down the template and interface using Word macros, templates and additional VB code. The issues with this are the following:

The solution tends to be either too general or too specific,
The cost and complexity of distribution makes them difficult to update,
Only certain versions of Word can be supported,
It can be difficult to support documents created against previous templates,
Training users to use the templates can be expensive.

Using ISO Schematron to express template specific validation rules allowed the ATO to have a light touch on the Word templates while still giving users the feedback necessary for their documents to conform to the production environment.

Rather than submit to a production queue for a human to assess, users could immediately see the issues and fix them prior to submission. This dramatically cut the lead times for production and allowed domain experts more time with their content. It also reduced the frustration the production staff experienced when having to fix the same problems more than once.

How to best process Word files is a discussion that will never end. What the ATO demonstrated is that following a standard template is a lot easier when users can see what they have done wrong. This is good for productivity and quality. It helps organizations to get the most out of their standard desktop and it allows users to control their documents without the cognitive burden of having to learn a lot of rules about technology that might only be a small part of their job.

Implementation

In the world of documents, Microsoft Word is ubiquitous and docx files will be the primary document format for a long time. However, no one would argue that Word is a large, sophisticated piece of software that requires a certain amount of time and dedication to master. These requirements can be more than people are willing to invest.

Schematron

Years of processing Microsoft Word-based documents into XML has proven that relying on users to understand and comply with template conventions is unrealistic and frustrating for everyone. Providing users with feedback about their compliance is a more productive approach that allows users to manage their own workflow at the same time as ensuring a baseline for document quality.

If users prefer to resolve issues at the end of their process, they can. When problems are addressed is a decision for the user. The only concern for production staff is that the problems are addressed.

The approach used at the ATO was to avoid using W3C XML Schema validation and instead use ISO Schematron. Schematron provides a more subtle, user-centric mechanism for expression compliance, including the ability to write contextual error and warning messages. This allows the validation process to provide users with more clarity and direction than traditional schema validation. This transforms validation from enforcement to educational and from compliance to instructional.

Key Features

The ATO Word validator is a highly secure, hosted service backed by OX application. Documents are processed, then all content is purged. It is available as a hosted solution or on-premises. The processing stages are:

Basic analysis and gross error checking,
Purge structurally irrelevant, or non-compliant formatting information such as typographical instructions or formatting overrides,
Generate a detailed count of document objects and non-compliant use of styles or content,
Generate a basic view of the converted output for tracing problems,
Process the document through a two-pass, round-trip conversion to ensure that no content is lost in the conversion,
A final count and compare of markup and content to confirm the integrity of the converted data.

After processing thousands of documents a month for several years, this solution has enabled the ATO to: