This codebase implements a neural guided learner, using the loreleai library found at: https://github.com/sebdumancic/loreleai
All self implemented files:
- Every file in dir /Search is used for the search.
- Every file in dir /utility is mainly used for formatting neural network inputs.
- Every file in dir /filereaders is used to read data from the given example and bk files
- /tests/solver_prolog_swipy_bktest.py implements a test, testing the evaluation capabilities of the SWI Prolog solver,
showing a possible bug (?)(fixed by changing the background knowledge file)
NOTE some other files of loreleai were also changes, these are:
- Several expansion hooks are added to the default ones in loreleai, found in /loreleai/learning/language_filtering.py
- The utitily methods used for the hooks in the previous point are found in /loreleai/learning/utilities.py