This project it's about experiments of the use of ILASP as a post-hoc method over black-box models such as Neural Network models, Support Vector Machines and K-Nearest neighbors. In the repository are present the files for the creation of the used dataset of recipes (raw data obtained by GialloZafferano website), several files about studies and experiments over the data and over the black-box models performances, and finally files about experiment with ILASP.
A full description of the dataset can be found here: https://zenodo.org/record/7265253#.Y2aZi3bMIuU
A full description of the experiments involved in this project can be found on the following articles:
1- D.Fossemò, F.Mignosi, L.Raggioli, M.Spezialetti, F.A.D'Asaro. Using Inductive Logic Programming to globally approximate Neural Networks for preference learning: challenges and preliminary results. Proceedings of BEWARE-22, co-located with AIxIA 2022, November 28-December 2, 2022, University of Udine, Udine, Italy. CEUR Workshop proceedings. Vol. 3319. 67:83. 2023. Url: https://ceur-ws.org/Vol-3319/paper7.pdf
2- F. A. D’Asaro, M. Spezialetti, L. Raggioli, S. Rossi, Towards an inductive logic programming approach for explaining black-box preference learning systems, Proceedings of the International Conference on Principles of Knowledge Representation and Reasoning 17 (2020) 855–859. DOI: https://doi.org/10.24963/kr.2020/88
3- D.Fossemò, F.Mignosi, M.Spezialetti, Analisi comparativa di approcci di machine learning per un sistema di preferenze di cibi. Three-year degree thesis (D.Fossemò). DOI: http://hdl.handle.net/20.500.12319/3727
4- M.D'Aviero F.Mignosi, M.Spezialetti, Apprendimento Induttivo di Answer Set Program per un sistema di preferenze di cibi. Three-year degree thesis (M.D'Aviero). DOI: http://hdl.handle.net/20.500.12319/2304