Summary
BMJ’s new product, called “Point Of Care” (PoC), provides medical knowledge to customers in the surgery or workplace (at the “point of care”). It needs to integrate with their existing Docato-based system of products as closely as possible to enrich content for syndication, sale and improved searchability for 'self serve' content
Key issues
BMJ needed to:
- Implement a new XML schema
- Link to existing clinical references and BMJ products
- Tag standard drug names on import
- Proofing of content to PDF
Solution description
The solution is designed to fit into BMJ's existing product framework as far as possible. Data is imported into Docato and split into content chunks for the major sections, references and images. On import, standard drug names are tagged and references to other PoC documents and existing clinical references are resolved. The output from the system is in the form of XML files for supply to content resellers.
This process is carried out using an automated file upload using a Java-based file/folder watcher. When a file is saved into a designated area, it is automatically loaded into the database. When the files are uploaded, all linked images and XML files are resolved. Missing items are imported as "abstract" resources until the actual file is loaded using the same mechanism.
As the files are loaded, specific file types are automatically sent on to HighWire to be put on the Web. Files that are in the system can be searched and edited using custom forms that allow for additional classification (topic codes) to be added and related material to be identified and linked.
The majority of the uploaded xml comprised articles that will be compiled and sent to HighWire as "virtual web publications".
There is a "shopping basket" that allows these publications to be generated from selected (searched) material that is then ordered as required and submitted with additional uploaded material (such as PDF versions of the topics).
HighWire generates an RSS feed from the data that it published. An additional task monitors the RSS feed and adds value to content that matches the feed items, providing information such as when and where the data is actually published.
The shopping basket is also used to generate other output types and has been designed to facilitate any further output types via configuration.
The Tools
- Java extensions to Docato
- XQuery to discover and update content
- Ant build files for publications
- XSLT data transformations to generate output files
Key Features / Benefits
- Implemented on existing Docato platform
- Links to existing content and content sales infrastructure
- Automated tagging of drug names enriches content
- Allows for custom creation of content using "virtual documents" and automated publishing
