We focus on the problem of inducing logic programs that explain models\nlearned by the support vector machine (SVM) algorithm. The top-down sequential\ncovering inductive logic programming (ILP) algorithms (e.g., FOIL) apply\nhill-climbing search using heuristics from information theory. A major issue\nwith this class of algorithms is getting stuck in a local optimum. In our new\napproach, however, the data-dependent hill-climbing search is replaced with a\nmodel-dependent search where a globally optimal SVM model is trained first,\nthen the algorithm looks into support vectors as the most influential data\npoints in the model, and induces a clause that would cover the support vector\nand points that are most similar to that support vector. Instead of defining a\nfixed hypothesis search space, our algorithm makes use of SHAP, an\nexample-specific interpreter in explainable AI, to determine a relevant set of\nfeatures. This approach yields an algorithm that captures SVM model's\nunderlying logic and outperforms %GG: the FOIL algorithm --> other ILP\nalgorithms other ILP algorithms in terms of the number of induced clauses and\nclassification evaluation metrics. This paper is under consideration for\npublication in the journal of "Theory and practice of logic programming".\n