Multivariate networks (MVNs) have become increasingly prevalent in contemporary research domains, spanning network security, biotechnology, and healthcare applications. These complex data structures are characterized by dual perspectives, as exemplified in social networks: an external perspective capturing internode relationships and an internal perspective encompassing node-specific attributes. Although existing visualization approaches predominantly employ dimensionality reduction (DR) algorithms to generate separate views for each perspective, this limitation impedes comprehensive network analysis.This thesis introduces a novel visualization approach that integrates both MVN perspectives into a unified low-dimensional representation. We present two distinct methodological modifications to the t-SNE algorithm, which enable simultaneous processing of both internal and external network characteristics. Furthermore, we propose an improved quality metric for evaluating DR algorithms that takes into account both MVN perspectives simultaneously. Our comparative analysis includes both a qualitative visual assessment and a quantitative evaluation of the proposed modifications, which contributes to the advancement of MVN visualization techniques.