Interaction in virtual worlds : a comparative study of the virtual assistants Siri and XiaoIce

Abstract

In the digital era, virtual assistants such as Siri, XiaoIce and have become an important bridge connecting human beings with the digital society. This paper explores in detail the technical support, development history and application prospects of these virtual assistants in the digital economy, social media and entertainment. It analyzes how virtual assistants can understand and respond to human needs and emotions using a variety of technologies such as natural language processing, affective computing, and speech recognition, and thus enrich the user experience. This study not only evaluates the differences in their performance in natural language processing, speech recognition and emotion computing tasks, but also reveals their respective technological strengths and weaknesses. The findings show that XiaoIce performs well in handling complex situations, especially in the area of affective computing, which is significantly better than Siri. furthermore, while Siri performs better in handling simple voice commands, it is limited in its ability to handle complex commands in high-noise environments. These findings point to the challenges of current AI technologies in terms of depth of understanding, speech recognition diversity, and emotional interactions, while providing directions for the development of technology optimization to enhance the usefulness and user satisfaction of virtual assistants.