Please note: The algorithm descriptions in English have been automatically translated. Errors may have been introduced in this process. For the original descriptions, go to the Dutch version of the Algorithm Register.
PRIMA Enrichment – Mission Archives
- Publication category
- Other algorithms
- Impact assessment
- DPIA
- Status
- In development
General information
Theme
Begin date
Contact information
Responsible use
Goal and impact
AI is being used to classify and enrich (with labels and metadata) documents and data in mission archives. The aim is to make them more searchable.
Considerations
The nature of the data makes it challenging to identify the relevant and correct versions of documents. The use of AI to organise and enrich the data is essential and serves as input for the search algorithm.
Human intervention
There is no automated decision-making involved. A human assesses the search results for relevance.
Risk management
Risk management forms part of the development process. Performance is monitored to ensure technical accuracy. The internal DPIA contains a comprehensive overview of the privacy risks.
Legal basis
The legal basis for the processing is Article 6(1)(e) of the GDPR: the processing is necessary for the performance of a task carried out in the public interest assigned to the controller. This relates to the legal obligation set out in Articles 2.4 and 4.4 of the Woo and Article 3 of the Archives Act 1995.
Elaboration on impact assessments
Impact assessment
- Data Protection Impact Assessment (DPIA)
- Pre-scan DPIA
Operations
Data
The data comprises all information generated during a mission and transferred to the Semi-static Information Management (SIB) archive.
Technical design
The applications in which the Ministry of Defence uses algorithms may be operationally sensitive or classified. The Ministry of Defence therefore only provides general information on the technical functioning of its algorithms. To enrich mission archives, the Ministry of Defence uses: Document parsing (to convert file types into a computer-readable format); NER (to recognise entities in text and create labels/metadata); LMM (for summarising and adding labels); various NLP techniques such as regex and modern NLP (to recognise document structure and type); embeddings (to support vector search and semantic search).