Public space notification

Municipality of Kampen

Residents can report through an online reporting system if something needs to be fixed or cleaned up in the public space (e.g. street or park). If the municipality is responsible for this, it will be dealt with.

Last change on 1st of May 2024, at 8:06 (CET) | Publication Standard 1.0

• Publication category: Other algorithms
• Impact assessment: Field not filled in.
• Status: In use

General information

Theme

• Public Order and Safety
• Space and Infrastructure
• Traffic

Begin date

2023-10

Contact information

[e-mail]

Responsible use

Goal and impact

Previously, reporters themselves had to choose to which main and subcategory their report best fitted, so that the report went to the right department of the municipality. Thanks to the algorithm, this is no longer necessary; the (sub)category is determined on the basis of the report text and the report is automatically classified. This saves the notifier (the resident) work and the employee too, since (usually) without any intermediate steps the notification goes straight to the team that has to deal with it.

Considerations

The reason for using the algorithm is the efficiency gains it makes.

Human intervention

Notifications do not always end up in the right category and the algorithm also does not always recognise which category the notification should go in (and then puts it in the Other category). In those cases, the notification still has to be put in the right category manually.

Risk management

There are few risks in this algorithm. Small risk is that the algorithm puts a report in the wrong category. In that case, a reporter may get follow-up questions that have nothing to do with the report made. In that case, the reporter is always given the option that the questions are not applicable, but could cause confusion.

Operations

Data

Notification texts are used for the algorithm. While making a report, the reporter can choose to leave a phone number or e-mail address on a new page, but this information is only used for providing feedback on status changes. This information is not kept longer than necessary for this purpose, nor is it used for the algorithm.

Technical design

The text of the notification is broken down into individual words. Each word from a notification is analysed to see how unique it is to that notification, set against the total collection of words ('TF-IDF' or 'term frequency-inverse document frequency'). As a result, a word like 'the' or 'thank you' gets a low weight and a word like 'rubbish' gets a higher weight. From that combination of words, logistic regression (a machine-learning technique) is then used to determine which category the report belongs to.