The Influential Node Identification task aims to measure which nodes in a network could spread information more efficiently. However, most works focus on homogeneous networks, whereas there are still a few approaches regarding heterogeneous networks. Existing works for identifying influential nodes in heterogeneous networks use metapath latent representations. This approach is limited because it depends on prior knowledge about the network to establish random-walk strategies based on vertex types. In contrast, Knowledge Graph Embeddings (KGE) can model heterogeneous networks differently than those obtained using random-walks by transforming the heterogeneous networks into a knowledge graph, thus not needing prior knowledge about the network schema of types. Therefore, this work aims to study the effectiveness of KGE methods for the Influential Node Identification task in heterogeneous networks. We use two existing scoring methods based on node representations adapted to operate with KGE. Moreover, we simulate using a well known epidemic diffusion model (SIR) and calculate correlations to evaluate those modified measures and their obtained node ranking in three datasets widely used in heterogeneous network tasks. Results show potential outcomes using the TransE model to identify influential individuals, obtaining Kendall’s coefficient close to 0.9 under the SIR model.