Lasswell’s “5W” Communication Model is a foundational theory in the field of communication studies, proposed by Harold Lasswell in 1948. This model deconstructs the communication process into five key elements: Who (Communicator), Says What (Message), In Which Channel (Medium), To Whom (Audience), and With What Effect (Effect). It facilitates a systematic analysis of communication behavior and its efficacy across different stages, and also provides substantial insights into the investigation of media message perception, affective evaluation, and interaction-based identification [2] . While applicable to traditional mass communication, the model is also adaptable to research on User-Generated Content (UGC) dissemination on digital platforms. Particularly on short video platforms, communicators—ranging from individual creators and corporate accounts to media outlets and government units—influence the cognition and attitudes of specific groups through diverse content packaging and channels. The “5W” model aids researchers in clarifying the chain of responsibility and value within communication acts, providing a structured analytical pathway for empirical research. This model offers a systematic framework for understanding and analyzing communication phenomena.

Social Network Analysis (SNA) is a quantitative methodology used to investigate social structures and their internal relationships. Based on graph theory, it is commonly employed to examine metrics such as relationship strength, centrality structures, and network density. In recent years, SNA has been widely applied in social media research and topic diffusion studies. In this research, SNA is primarily used for textual analysis of video content [3] . Regarding data collection, this study employed Python scripts to scrape data from Bilibili. Using the search keyword “Shanghai Metaverse”, 678 video entries published between January 1, 2019, and May 22, 2025, were collected. The dataset encompasses nine fields: video title, publication time, view count, like count, favorite count, coin contribution count, description, comment content, and publisher information. Furthermore, the GooSeeker [4] text processing tool was utilized for word segmentation and the construction of a keyword co-occurrence matrix to facilitate subsequent content network analysis.

The first Metaverse video identified, titled “Metaverse Shanghai City VR Virtual Reality Virtual City Virtual Community Virtual Showroom”, was published by the company “BlueView Tech Metaverse” on March 12, 2019. From March 12, 2019, to May 22, 2025, the annual distribution of videos retrieved using the keyword “Shanghai Metaverse” is as follows: 1 (2019), 1 (2020), 83 (2021), 305 (2022), 192 (2023), 79 (2024), and 17 (2025-partial year). This trajectory indicates that Metaverse communication experienced phases of inception, growth, and subsequent decline. Analysis identified five primary types of communicators: Public Service Institutions, Media, Education, Companies, and Individuals. Individual communicators constituted the largest proportion at 62%, with some potentially acting as proxies for companies (disseminating company products/services under a personal identity). The remaining four categories each held relatively smaller but comparable shares, collectively accounting for 38%. Table 1 details the communicator types, representative examples, and illustrative content examples.

The titles and descriptions of all 678 videos were integrated to form the content corpus. Using the GooSeeker text analysis tool, the corpus underwent word segmentation, automatically generating a co-word matrix. Subsequent visualization analysis of this matrix via SNA software, applying filters for word frequency > 3 and co-occurrence links > 2, yielded the Shanghai Metaverse video content co-word network ( Figure 1 ). Link thickness corresponds to co-occurrence frequency. The analysis reveals that the primary foci within the videos are Digital Twins and Digital Collectibles (NFTs), Metaverse Cultural Tourism, Libraries, and the relationship between the Metaverse and Artificial Intelligence.

To further capture nuanced narratives, we propose a supplemental qualitative analysis of the top 50 viewed videos (top 7.4% by view count). Preliminary observations suggest three recurrent narrative frameworks: 1) Techno-utopianism (emphasizing inevitable technological progress); 2) Critical pragmatism (discussing hardware/ethical constraints); 3) Commercial instrumentalism (prioritizing product promotion). Future research should conduct systematic coding to quantify these patterns and identify gaps in socioeconomic governance discourse.

All video data in this study were collected from Bilibili. This platform, characterized by its strong community attributes, “danmu” (bullet comment) culture, and young user demographics, has become a significant channel for disseminating virtual reality and technology content in recent years. Bilibili’s algorithm-driven recommendation system and culture of secondary creation (remixes, parodies, etc.) also offer greater potential for the diffusion of Metaverse content and avenues for creative expression.

Compared to short video platforms like Douyin (TikTok) and Kuaishou, which predominantly feature fragmented entertainment content, Bilibili’s videos tend to be more knowledge-oriented and structured [5] . This characteristic is more conducive to presenting and explaining the complex issues surrounding the Metaverse. However, it is also noted that technology videos on Bilibili face a certain “dimensional barrier” to dissemination: community layers are heavily solidified, making it difficult for content to break out beyond specific interest groups and achieve broader reach.

Furthermore, the study found that some Bilibili Metaverse videos are cross-posted [6] to other social platforms like Weibo and Xiaohongshu (rednote), enabling multi-channel coordinated dissemination. Future research could further explore the impact of “multi-platform dissemination mechanisms” on Metaverse communication effectiveness to enhance the integrated analysis of dissemination pathways.

Audience analysis examines audience characteristics and feedback to understand their needs and reactions. Comments posted below videos provide the primary data source. Analysis of the collected 25,764 comments (averaging 38 comments per video) using GooSeeker for sentiment analysis revealed the following distribution: Positive comments constituted 46%, neutral comments 48%, and negative comments 6%. Positive comments generally expressed support for Metaverse videos but were often simplistic (e.g., “冲冲冲”—“Charge ahead!”, “Wow”), lacking detailed or substantive critique. Neutral comments primarily involved discussions tangential to the video topics, such as the performance and price of VR equipment. Negative comments mainly questioned whether the videos depicted a “true” Metaverse, with gaming-related content being the most contentious aspect.

Effect analysis evaluates the impact of communication on audience cognition, attitudes, and behaviors, measuring communication effectiveness. Video metrics such as view count, likes, favorites, and coin contributions serve as proxies for assessing effect. For the 678 collected Shanghai Metaverse videos, the aggregate metrics are: Total Views: 42.6632 million, Total Favorites: 1.4892 million, Total Likes: 1.5623 million, Total Bookmarks: 0.8654 million, Total Coin Contributions: 87,000. Comparing these figures with a control group of 702 “Shanghai Artificial Intelligence” videos (retrieved via Bilibili search using “Shanghai Artificial Intelligence” as keyword during the same period: Jan 2019-May 2025), the dissemination effectiveness for Shanghai Metaverse content is significantly weaker. Specifically, the view count of just the top 30 “Shanghai AI” videos by popularity exceeded the total view count of all “Shanghai Metaverse” videos. Deeper underlying reasons likely include: AI technology demonstrating stronger real-world applicability; the Metaverse facing high hardware barriers and limited experiential accessibility; and declining public interest following the initial hype cycle. The observed pattern of high view counts but low comment counts suggests a “high bounce rate”, indicating users browse the content without engaging deeply.

To gain further insight into the dissemination dynamics of Shanghai Metaverse videos on Bilibili, this study adopted a “lifecycle” perspective, statistically analyzing publication times and subsequent interaction curves. The results indicate that the majority of a video’s views and likes are concentrated within the first 7 days after publication, followed by a rapid decline. This pattern confirms that “instantaneous dissemination” remains the dominant mode on short video platforms.

A few videos published by institutions exhibited viewership peaks coinciding with “Metaverse policy announcements” or “major conferences,” demonstrating an “event-driven” characteristic. This suggests that content creators should strategically time releases to align with significant technology news, industry policies, or festival themes to enhance the visibility and topic longevity of Metaverse content.

Additionally, a small number of videos experienced “secondary dissemination” (e.g., through re-uploads by the original UP主, reposts, or edited clips) approximately one month after initial publication. However, such instances were rare, indicating that Bilibili currently lacks a well-developed “long-tail dissemination mechanism” specifically for Metaverse videos.