Knowledge graphs are incomplete by nature, with only a limited number of
observed facts from the world knowledge being represented as structured
relations between entities. To partly address this issue, an important task in
statistical relational learning is that of link prediction or