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.
BEAR
- Publication category
- Impactful algorithms
- Impact assessment
- Field not filled in.
- Status
- In use
General information
Theme
Begin date
Contact information
Responsible use
Goal and impact
The purpose of BEAR is to make important information visible faster. The system helps intelligence analysts to quickly find what is of interest in large quantities of documents.
BEAR can be used when it is necessary to form a (personal) picture of the information already known to the police. This is important in arrests, for example. An arrest team needs to know whether a suspect has used a weapon before. In crisis intervention, understanding someone's mental state helps to tailor the approach accordingly. In the detection of fugitives, background information supports the investigation process.
As a result, analysts spend less time reading documents that are not needed. More time is left to properly assess and analyse the important information. This is especially useful when working under time pressure. This reduces the chances of missing important information and improves the final assessment.
Considerations
Now analysts search documents themselves with search terms. This does not include the context of a sentence, and spelling errors can affect the results. Summarising documents with a language model was also considered. This was not chosen, as it is precisely the details that are important for analysts. Therefore, a solution was chosen where the original content remains and documents are displayed in order of importance.
The big advantage of BEAR is that analysts can find and read important information faster. This allows them to make faster and better decisions. This can contribute to the safety of citizens and police officers. It allows an arrest team to better prepare and tailor deployment to the situation. This is especially important in rapid deployment where there is little preparation time.
With the use of BEAR, it is still possible for information to be missed or (also) shown that is not important. BEAR only ensures that documents are presented in a logical order. The documents must still be read and assessed by the analysts themselves. The analyst remains responsible for their own explanations and choices. The analyst must also be able to explain why information has been assessed in a particular way. BEAR supports the analyst and does not provide a ready-made answer. This reduces the chance of errors in the algorithm distorting the result.
Human intervention
The analyst must always form their own judgement on the information found. The algorithm helps by pre-ordering documents. This allows the analyst to see more quickly which information might be of interest.
The system only gives a clue about the topic. Human insight remains necessary to understand and assess the content. All information remains fully visible to the user.
Risk management
The design deliberately chose to keep humans at the centre. The algorithm does not provide answers to questions. It only helps the user find answers himself. In addition, it technically checks whether the model works as intended. In this way, the system remains reliable and is used appropriately.
Legal basis
Articles 8 and 9 in conjunction with Articles 11 and 13 Wpg. Maintain public order and search for suspects.
Operations
Data
Data were compiled manually. Sentences were classified by people by topic. In doing so, a conscious decision was made not to use personal data.
Technical design
The algorithm determines whether a sentence is about a particular topic. The text of a sentence is first converted into numbers. This is done using a text model.
Then the system learns which patterns belong to which topic. Based on this, the system can estimate which topic they fit in new sentences. It can also happen that a sentence does not belong to any topic.
Similar algorithm descriptions
- The Operational Readiness Dashboard provides insight into the current operational readiness and deployability of the armed forces. Specifically, the dashboard provides insight into personnel readiness, materiel readiness and exercise readiness per unit. This dashboard runs in BI/VEST, which stands for Business Intelligence/Improved Steering Information.Last change on 28th of November 2024, at 12:30 (CET) | Publication Standard 1.0
- Publication category
- Other algorithms
- Impact assessment
- DPIA
- Status
- In use
- This algorithm assigns a priority category based on data recorded in the environmental department's case system.Last change on 3rd of December 2024, at 13:23 (CET) | Publication Standard 1.0
- Publication category
- Impactful algorithms
- Impact assessment
- Field not filled in.
- Status
- In use
- The Summarizer allows ACM employees to summarise documents, especially reports and decisions, using an internal language model (currently GPT-5-mini in the ACM Azure environment). Confidential documents may also be summarised because of privacy and information security agreements with the vendor.Last change on 24th of February 2026, at 19:34 (CET) | Publication Standard 1.0
- Publication category
- Other algorithms
- Impact assessment
- IAMA, DPIA
- Status
- In use
- The algorithm calculates which absence reports made by schools can be assigned to which compulsory attendance officer.Last change on 5th of July 2024, at 10:32 (CET) | Publication Standard 1.0
- Publication category
- Other algorithms
- Impact assessment
- Field not filled in.
- Status
- In use
- This algorithm helps users of BuitenBeter determine what subject a report is about. This is an application that allows reports to be submitted about public spaces.Last change on 16th of January 2026, at 9:29 (CET) | Publication Standard 1.0
- Publication category
- Other algorithms
- Impact assessment
- Field not filled in.
- Status
- In use