The Australian Human Rights Commission (the Commission) is Australia’s national human rights institution, an independent statutory organization in the Attorney-General’s portfolio that promotes and protects human rights. Although a small agency, AHRC has a reputation for early strategic adoption of new technologies. In recent years AHRC has been a leader in government with their migration to Microsoft Office 365 and Azure.
According to the National Archives of Australia (NAA), the Commission is the current thought leader of records management in the Australian Government. Through a corporate partnership with RecordPoint, the Commission implemented an electronic document and records management system (EDRMS) on SharePoint with RecordPoint Records365, utilizing RecordPoint’s AI and machine learning technologies to classify records without the need for staff input.
Prior to this new EDRMS solution, in production since February 2019, the Commission was drowning in a sea of duplicates, tangled in nested folders and perplexed by lost documents. Funding shortfalls and other challenges saw the Commission unable to implement an EDRMS solution that was viable.
Researching options, the Commission headed by Ron McLay, Chief Information Officer and Ryan McConville, Information Manager, incorporated the Department of Finance’s study into the failings of the traditional EDRMS. In particular, the report suggested that records management should be automated, rather than a being a manual task for public servants. Inspired by the report, the Commission set out to implement a fully automated EDRMS, using artificial intelligence (AI) and machine learning. This would form the basis of RADICAL – Record And Document Innovation & Capture – Artificial Learning.
An EDRMS implementation with AI addresses key problems faced in running an effective EDRMS. AI reduces the scope for human error, while increasing the volume, accuracy and consistency of records classification. The simple user interface has driven high user uptake and is seen by staff as a useful tool rather than a burden.
The Commission avoided customization and add-ons for RecordPoint and SharePoint, focusing on configuration instead. A common problem agencies have experienced is in the customization and use of third-party add-ins to suit existing or outdated business processes. This often resulted in systems that were difficult to use, inefficient and unreliable, and hard to upgrade with user uptake suffering accordingly.
Harnessing the native functionality of RecordPoint and SharePoint translated to improved business processes. The Commission also incorporated simple navigation for easy browsing of records, supported by a powerful search feature in Records365. The RADICAL solution has been in production since February 2019 and is currently in staging for deployment across the organization
The Commission’s approach was ‘configuration over customization’ as recommended by the DTA, focused on human-centered design. Staff were consulted extensively on current needs and pain points. When possible, native Records365 and SharePoint functionality was preserved, limiting the need for end user training and burdensome change management.
When planning RADICAL, a key goal was to remove records management decisions from staff and allow them to focus on their core work. RADICAL needed to provide ‘transparent records management’ and limit the potential for inaccurate or inconsistent classification.
Traditionally, the classification process has been performed manually by records officers. The manual element of classification can be time consuming, can lead to inaccuracy and can be disruptive to staff. Previous methodologies to automate records classification uses rules trees that classify records based on their metadata and saved location. However, rules trees need to be built and maintained by experienced records officers and rely on end users to apply accurate metadata and save to specific locations.
Leveraging AI in this process solves many of these problems by combining a minimal rules tree with a machine learning model. If a record cannot be categorized by a rule, the machine learning model classifies the record based on its contents. This system eliminates the need to maintain complex rules trees, the reliance on metadata and record location.
The RADICAL project team worked with RecordPoint’s AI developers to create a statistical model that can classify records against AFDA Express and the Commission’s agency-specific records disposal authority.
The statistical model is developed by taking a set of records that have been manually classified and applying Natural Language Processing techniques to normalize the document content into vectors. The model is then trained using algorithms.
After an initial training period, the RADICAL statistical model can categorize individual records with an accuracy of 80%. The Commission expects this accuracy will increase over time. RADICAL also re-categorizes records each time they are edited, ensuring the classification is always current.
Although the machine learning model will initially work in conjunction with a rules tree, as the accuracy of the model increases the rules will be gradually removed and the Commission will rely solely on machine learning to manage their corporate records.
The implementation phase involved:
- performing a detailed analysis of existing systems and record holdings
- developing a new information governance framework and agency-specific records disposal authority
- developing and implementing a records migration strategy
- extensive collaboration and testing with the RecordPoint AI developers
- developing and testing of the SharePoint platform that Records365 manages
- working with a specialist change management facilitator
- training end users and providing ongoing support on go live
- implementing an agreed security model
- gaining agency-wide approval for the system including the business rules and for use of the machine learning algorithms
- training the machine learning algorithms
- implementing and rolling out the system
As most Australian Government agencies share the same records management requirements, the Commission feels that the machine learning model provided by RecordPoint and used by RADICAL is a ‘genuine game-changer’ and will allow other agencies to experience equivalent ‘gains in efficiency, productivity and cost reductions.’ The Commission sees themselves as trailblazers in government for the use of AI in records management and are excited to share their experiences as the current thought leaders of Australian Government records management.
RADICAL provides multiple, tangible benefits to the Commission such as:
- automated records management that is accurate, consistent and compliant
- compliance with DC2020
- document versioning, which reduces duplication
- enhanced collaboration and sharing
- streamlined handling of Freedom of Information (FOI) requests
- Power BI reports for senior executives
- reduction in staff time spent on records management
- effective and efficient records search and retrieval
- real-time video transcription
- automated image cataloguing
RADICAL has had a positive impact on the Commission and its stakeholders by:
- delivering upon the objective of ‘transparent records management’
- increasing the accuracy and compliance of information management practices by reducing the scope for human error
- reducing the time and costs associated with responding to FOI requests through improved search and retrieval
- gradually reducing physical storage costs, currently averaging $17,000AUD per annum
- reducing digital storage costs
- increasing collaboration between Commission business units through shared document libraries and the establishment of an ‘open by default’ information access policy, where access to records is restricted only to protect personal privacy or sensitive information
- improving business processes through electronic workflows, document versioning and automated metadata tagging
- minimizing the impact of potential data breaches through regular scheduled records disposal
Initial estimates by the Commission suggest that staff using RADICAL are seeing at least a 5% increase in productivity. Additionally, the accuracy of capture and classification by the algorithms is improving, and by estimates ‘it already exceeds the accuracy of our manual classification.’
Lastly, the Commission showcases that a technologically advanced solution can be implemented without significant costs. ‘We estimate that a traditional EDRMS would have cost the Commission 3 or 4 times as much as RADICAL.’