A total of 16 sites, with its own editorial staf and CMS were to be transformed into a common website, Government.nl based on Hippo.
Automated content migration of all content to Hippo CMS, including enrichment and cleanup.
Successful, fragmented migration to Rijksoverheid.nl with cleaned and valuable content.
Government.nl is the main English-language news site for the Dutch government. It publishes news and information on one single site. The first version of Rijksoverheid.nl went live in 2010 and replaced 14 individual department websites, and the sites Regering.nl and Postbus51.nl. Prior to the migration each department had its own website with its own editors, content management system and processes. The objective was to create one common website for managing information where the content of the existing websites is housed. For this, Xillio was asked to help.
The old department specific websites used among others Green Valley, I-Park, Viadesk, Tridion and Iprox CMSes. For the new all-in website the open source CMS Hippo was chosen. The choice was made to run the migration of the 16 websites concurrently with the large releases from Rijksoverheid.nl. Previous to the migrations a ten day pilot was performed wherein the content of MinVWS.nl and Postbus51.nl was delivered in a predefined XML format. Additionally XSLTs were developed that converted the XML to HTML-pages, which made the XML files better readable for the testers.
After the pilot an inventory began in which the content types of each department were fully described. For each department and content type the metadata and fields were documented and several overviews and matrices were created on department, content type, metadata and field level where the similarities and differences became visible. In addition, a mapping was created for each department on content type and field level.
Where applicable, mapping rules have been added. These range from conversion lists for the automatic linking to subjects, departments and ministers up to the formulation of rules for assigning the required metadata fields and conversion to the correct format.
Enrichment and cleaning
During the migration the content was enriched with the metadata that had been collected and converted to the strict format of Rijksoverheid.nl and Hippo CMS. The department websites’ content was converted to XML. This XML was then sent to the Hippo importer that was developed by Xillio during the project. The importer collects the XML and imports it into Hippo CMS.
The websites usually contained similar content types, but the design and the presentation differ immensely. During migration the different forms were converted to the government’s newly defined structure for parliamentary documents. Besides that other documents like Word and Excel were converted to PDF. The migrated content was cleaned in different ways to have it conform to the government standards.
The content on the department websites often consisted of one HTML block. During the migration the content was split into different blocks for paragraphs, pictures and video. This was delivered to the importer as separate blocks. Not only was the content of the HTML cleaned, also the content set itself. Several rules were applied based on, for instance, document date, context or content type to clean the content set.
The first migration on March 31 2010 was combined with Rijksoverheid.nl going live, but the following migrations were performed directly into the live environment. That meant there was no room for errors.
Two types of internal links were found during the migration; the internal links within a departmental website and the previously external links between departments. The links have been converted and provided with additional information for the Hippo importer to be able to reconstruct these links during migration.
The documents and images have been encoded so they can be send within the XML SOAP message. From the document and image context and from the documents and images themselves the metadata is extracted and converted to the right format. The videos are not saved in the CMS of Rijksoverheid but are available at Rijksoverheidsvideo. During migration the video metadata was recognized and converted to the correct format.
The content is migrated straight into and published on the live environment of government.nl in scheduled, automated manner.