In the era of rapid technological development, the integration of artificial intelligence and virtual reality technology has gradually become a research hotspot. Interpersonal interaction, as an important part of virtual reality, also presents opportunities and challenges. The study builds the virtual interpersonal interaction model using the emotion model and hierarchical emotion as the learning algorithm after exploring the key technology of virtual interpersonal interaction based on the UGC model. Finally, the complex interpersonal interaction in virtual reality is simulated to examine the use of virtual community users and the factors affecting interpersonal interaction in virtual communities. According to the results, the percentage of female users of virtual communities is 14.28 percentage points higher than that of male users. The majority of people aged 18-40 who use virtual communities account for 88.01% of the population. In terms of education distribution, the group of bachelor’s degree holders is the group that uses virtual communities the most frequently, accounting for 66.52%. The study concluded that 8 out of the 11 hypotheses it proposed were valid. Among them, hypothesis 3 and hypothesis 11 were supported by 0.01 significance, and hypothesis 9 had a significance of 0.017, which reached 0.05 significance level.