This coding challenge is to be done for the position as NLP Engineer for the joint project of the Robert Koch Institute and the World Health Organization Hub on public health intelligence. The deadline is 2024-01-27, 23:59 CET.

We made two queries to our ElasticSearch database. The queries returned information on potentially relevant articles about public health events from two groups with different surveillance objectives. The files are named data_query_from_9367.json and data_query_from_9578.json. Your task is to build a classifier that can identify whether an article belongs to query 9367 or 9578 using only text data (title, description, or fullText). Do not assume that data is clean.

Additionally, you should demonstrate your skills by analyzing the data in one more way. This additional analysis should be related to natural language processing, however, there is no other restriction.

• Aim to spend no more than 2-3 hours on the aforementioned tasks.
• Create a private repository for your work on GitHub and invite aauss to view the repository:
  • Make your work reproducible. We will not contact you if something is unclear.
  • Document your work well but also be concise.
  • Your work should run on a basic laptop.
  • No paid services should be required to run your code.
  • Make a release on GitHub before the deadline. Then send me an invite. You should not make any commits after the deadline.
• Follow best practices of writing code.
• Follow best practices in machine learning and natural language processing.
• Present your work well. This can be done in a notebook or similar.
• Discuss results, bugs, and confusion. Be critical.
• Accurate performance on the dataset has the least priority.
• You will not be able to do all the tasks equally well in this time. Plan your time well.
• You can use any package or programming language to complete your tasks. There is a preference for Python but you can use any language you feel most comfortable with for this task.

We will use the aforementioned evaluation criteria to mark your work. Furthermore, we will ask you specific questions about your code and approach. Lastly, we will try to challenge certain aspects of your work. This may include mistakes or missed opportunities for improvement.