Matrač project

[Stanoja]

• wikidata SPARQL query adjustments (list of known exclusions)
• added fetching from related articles and keyword extraction
• added local llm execution in order to categorize items
• TECH: added Makefile with regular commands
• TECH: added root requirements.txt that also installs dev dependencies
• TECH: Fixed Github test action not to run twice when pushing on open PR (this PR for example)

TODOs:

• Test and add the feedback mechanism for categorization ✅
• Fix for rate limiting on related article fetch ✅
• Make both the keyword extractor and article fetch lazy-loaded and non-blockers for the process (log the error and resume) ✅
• Update README.MDs ✅
• Push rest of the tech improvements
  • Fix the build failure due to dependencies ✅
  • Push environment management ✅
  • Ignored venv for flake8 ✅