This project updates the Brussels Network for Science Promotion (STEAM) catalog by leveraging web scraping techniques to identify new STEAM actors. A Python-based script systematically extracts and analyzes external weblinks from 95 existing STEAM organizations’ websites, using a keyword-based classification system to detect relevant actors and expand the network.

The challenge

As part of the STEAM action plan for Brussels, this project focuses on updating the catalog of the Brussels Network for Science Promotion, which maps various STEAM actors and their initiatives. The goal is to provide an up-to-date overview of organizations engaged in science outreach and education, facilitating collaboration and strategic planning for strengthening science and technology engagement in the region.

Approach & result

To achieve this, a web scraping approach was employed to identify new STEAM actors beyond those already included in the 2018 catalog. Since STEAM actors often reference each other on their websites, the study leveraged advanced Python-based web crawling techniques to explore the online network of science promotion in Brussels. The process involved two key steps:

1. Identifying New STEAM Actors – The web scraping script navigated through the websites of the 95 STEAM organizations in the existing catalog, collecting external weblinks to other potentially relevant actors.
   
2. Classifying Websites as STEAM-related – The collected links were analyzed to determine their relevance based on a predefined set of STEAM-related keywords. These keywords were categorized into five themes: scientific topics (e.g., physics, chemistry, technology), type of activities (e.g., workshops, conferences, training), target audience (e.g., students, teachers), purpose (e.g., learning, exploration), and Brussels municipalities. To ensure comprehensive coverage, the keyword database was developed in Dutch, French, and English. A website was classified as STEAM-related if it contained at least one relevant keyword from each category.

This methodology identified 323 new potential STEAM actors, which were manually reviewed to ensure accuracy. By updating the catalog with this refined list, the project strengthens the visibility and connectivity of Brussels’ science promotion ecosystem, supporting efforts to make science more accessible and engaging for a broad audience.