With the rapid development of artificial intelligence, many applications have emerged, sparking heated discussions across various social sectors, particularly in education. This study aims to delve into the impact of pedagogical agents on learning through a combined use of VOSviewer and CitNetExplorer. Specifically, VOSviewer will facilitate a quantitative analysis, identifying top-cited authors, organizations, countries, co-cited authors, and frequently occurring keywords within the citation network. Meanwhile, CitNetExplorer will undertake a qualitative exploration, anticipated to reveal how visual stereotypes and emotional factors of pedagogical agents substantially impact learners' outcomes. This study lays the groundwork for further investigations into optimizing pedagogical agents' parameters to align with educational goals and promote their integration into instructional practices.