BLINK VS rules

I recently used a machine learning model from Github to predict different sets of documents. But what I am trying to do now is create a scoring script, which takes the same testing datasets but with different models. The output should have precision,recall, f1, etc. (something like that)

there is also BLINK by Facebook

https://github.com/facebookresearch/BLINK

Is it possible: yes.

Can you write your own home-grown rules engine in Ruby: yes.

Can you use off-the-shelf gems: yes. Here's a few I poked around in my previous explorations into Ruby Rules Engines:

* durable rules - https://github.com/jruizgit/rules?tab=readme-ov-file#ruby

* wongi - https://github.com/ulfurinn/wongi-engine

* rules - https://github.com/azach/rules

* ruleby - https://github.com/Ruleby/ruleby

  - bonus: video of original ruleby author explaining rules engines: https://www.youtube.com/watch?v=qMh2RDL6aBM

Currently https://github.com/nilp0inter/experta but https://github.com/noxdafox/clipspy seems nice, I just shied away from using it due to uneasiness about FFI and debugging, even though the original CLIPS is still awesome and has a very interesting manual.

There's also https://github.com/jruizgit/rules but haven't tried it yet.