A new statistic based on the sentiment of Flemish Tweets
Projects and research
The Data Science Hub produces and conducts research on new statistics about Flanders using new data sources and data science techniques. The Hub’s projects may directly output a statistic, or may aim to answer specific questions about an experimental statistic.
The new statistics provide value for policymakers, organisations, knowledge institutions and citizens.
The Data Science Hub creates and conducts research on new statistics:
- On request: when entities within the Flemish government need new or better statistics on a specific policy topic.
- On its own initiative: when an opportunity is identified where new data sources or data science techniques can lead to new statistics.
- In the context of academic research: the Data Science Hub collaborates with knowledge institutions in researching new methods for producing new and improved statistics.
If you need new/better statistics, you can always contact us for an exploratory meeting with one of the experts at the Data Science Hub.
Ongoing projects
Better company statistics based on natural language processing and web scraping.
This project investigates the possibility of predicting the relocation probability of citizens using register data.
Finished projects
A machine learning model that automatically categorizes companies as innovative or not based on the text scraped from their websites.
In this project, different categories of sentiment models are compared based on their performance categorising Flemish Tweets.
Recording of the SV-seminar "Data science for public statistics - research results academic collaboration"