Hotspots and Trends in Domestic Information Cocoon Research

Liu Xiaowen, Bai Rujin
(School of Journalism and Communication, Lanzhou University, Lanzhou, Gansu 730000)

Abstract

[Purpose] This study examines the landscape of domestic information cocoon research in recent years, exploring its hotspots and trends to provide a reference for related research. [Method] Using the visualization analysis software CiteSpace, we conducted keyword co-occurrence analysis and cluster mapping of literature related to information cocoons. [Results] The study reveals that recent domestic information cocoon research revolves around two core themes: "technology" and "people." However, it also identifies problems such as a singular research perspective and insufficient development of localized theories. [Conclusion] To address these issues, this paper proposes two developmental paths: first, breaking free from the constraints of Western academia, and second, transcending the research perspective of technological determinism.

Keywords: information cocoon; visualization mapping; CiteSpace; co-citation analysis; co-occurrence analysis
Classification Code: G203
Document Code: A
Article ID: 1671-0134(2025)04-34-05
DOI: 10.19483/j.cnki.11-4653/n.2025.04.006
Citation Format: Liu Xiaowen, Bai Rujin. Hotspots and Trends in Domestic Information Cocoon Research [J]. China Media Technology, 2025, 32(4): 34-37, 73.

1. Data Sources and Research Methods

This study employs bibliometric methods and utilizes CiteSpace 6.3.R1 software to conduct quantitative analysis of Chinese-language literature in the information cocoon domain. Through visualization-based screening and organization of bibliographic data, we generated keyword co-occurrence, keyword clustering, and keyword timeline maps to intuitively present current research hotspots, cutting-edge trends, and provide a solid foundation for the development of information cocoon research in China.

The data for this study were sourced from the China National Knowledge Infrastructure (CNKI) database. To ensure literature quality, we used "information cocoon" as the subject term, focused on the "journalism and communication" discipline, and conducted an advanced search within CSSCI-indexed journals. After removing invalid literature, we obtained 248 valid articles. The data collection date was August 27, 2024. This research is supported by the Gansu Provincial Soft Science Project "Paths and Strategies for Developing New Scenarios of Digital Cultural Consumption in Gansu" (Project No. 22JR11RA093).

2. Research Data Analysis

2.1 Annual Distribution of Research Output

For any research field, the number of published core papers serves as a crucial indicator for measuring research popularity and frontier trends. The overall trend chart of core literature publication volume can clearly demonstrate the research hotspots and trends in the field. As shown in [FIGURE:1], the earliest recorded output in China's information cocoon research domain appeared in 2010. The development of social media platforms such as Weibo and WeChat gave rise to phenomena like information cocoons, prompting scholars to begin paying attention to this area. From 2010 to 2017, information cocoon research was in its initial stage, with studies primarily focusing on defining the concept and analyzing its social implications, particularly concerning its negative effects on democratic society and proposing corresponding solutions. From 2018 to 2021, information cocoon research showed an overall upward trend, concentrating on information cocoon phenomena following the emergence of new media and artificial intelligence technologies, as well as the resulting problems and countermeasures. From 2022 to the present, research has gradually become more theoretical, with empirical studies on the generation mechanisms of information cocoons becoming popular and research beginning to show diversification.

2.2 Research Hotspots and Trends

In CiteSpace, we set "TimeSlicing" to 2010–2024, with "Years Per Slice" set to 1, using "keywords" as nodes to generate the keyword co-occurrence map shown in [FIGURE:2]. Larger nodes in the map indicate higher keyword frequency. The visualization map intuitively reflects the top ten hotspot keywords in information cocoon research: information cocoon, algorithm, artificial intelligence, algorithmic recommendation, echo chamber, big data, grounded theory, information literacy, influencing factors, and algorithmic ethics.

Building upon the keyword co-occurrence map, we conducted cluster analysis using CiteSpace software. The keyword clustering map is shown in [FIGURE:3], revealing a Modularity Q value of 0.6218 and a Weighted Mean Silhouette S value of 0.9662, indicating a significant keyword network structure, good clustering effect, and reasonable clustering results. Research in the information cocoon domain primarily consists of nine cluster segments.

Based on the overall publication trend chart, keyword co-occurrence map, keyword clustering map, and combined with analysis of relevant literature, we can summarize two major trends in China's information cocoon research: first, research on the generation mechanisms of information cocoons centered on technology such as "algorithm" and "artificial intelligence"; and second, research on the impact mechanisms of information cocoons centered on people such as "college students" and "group polarization." Notably, the college student group has emerged as an important subject for information cocoon research.

To analyze the dynamic evolution of this field, the keyword timeline map serves as an indispensable reference. Using CiteSpace software, we generated the keyword timeline map shown in [FIGURE:4], which roughly divides China's information cocoon research into three stages: the embryonic period (2010–2015), the budding period (2016–2019), and the golden period (2020–present).

During the embryonic period, only three research outputs addressed the concept of information cocoons as the concept was newly introduced to China's journalism and communication academia, with research remaining at the level of partial concepts of information cocoons involved in group polarization studies. In the budding period, with the emergence of algorithms and AI technologies, information cocoon phenomena became increasingly visible, yielding abundant research results focusing on technical-level generation mechanisms. Since 2020, academic research on information cocoons has gradually deepened, producing a large number of high-quality research outputs, with studies beginning to shift toward people-centered impact mechanism research, particularly regarding the college student group.

However, high frequency does not necessarily indicate strong betweenness centrality. Betweenness centrality measures the importance of nodes in the network, with higher values representing more important literature. As shown in [TABLE:1], information cocoon, artificial intelligence, algorithm, big data, and algorithmic ethics are key focus areas in this domain.

3. Technology and People: Two Cores of Information Cocoon Research

3.1 Technology-Centered Research on Information Cocoon Generation Mechanisms

The technological perspective has consistently permeated information cocoon research, particularly following the emergence of new technologies. Research on generation mechanisms in these technical contexts and comparative studies with previous technologies has become a major trend.

First, research on the generation process and negative effects of information cocoons in the context of Internet technology during the Web 2.0 era, with platform-based media such as Toutiao and Tencent News as primary examples. Li Zhenhui and Zhu Tingting conducted a dialectical analysis of the advantages and disadvantages of personalized recommendation technology, briefly analyzing the process of information cocoon formation and the fact that users are imprisoned without realizing it. Wang Qian used quantitative methods to survey users on the personalized recommendation platform Toutiao, laying a preliminary foundation for quantitative research on information cocoon generation mechanisms.

Second, research on information cocoon generation mechanisms on content intelligent distribution platforms in the context of big data and artificial intelligence. Wang Yicheng et al. analyzed the information flow methods on content intelligent distribution platforms through qualitative research and subsequently proposed an "information cocoon formation mechanism" model. Tang Lijia and Zhao Zhiqi analyzed the information cocoon generation mechanism on the Douyin platform from a big data perspective and explained its negative effects. Shen Nan used media form development as a division basis and analyzed the formation process of information cocoons from the dimension of audience psychological cognition. Wang Yaqian discovered through qualitative research methods that users jointly create information cocoons through personified expression and emotional interaction. Li Longfei and others, based on information ecology theory, explored the generation mechanism of information cocoon effects from the perspectives of information users, information, information environment, and information technology. Notably, some scholars verified and compared ten network media and found that whether the communication structure is horizontal or vertical, and whether user connections are open or closed, constitute two key mechanisms determining whether network media bring about "information cocoons." With the emergence of generative AI technology, scholars have begun researching this new technology, and China's research on information cocoon generation mechanisms has initially woven a research network centered on media, covering the Internet, big data, algorithms, and artificial intelligence.

3.2 People-Centered Research on Information Cocoon Impact Mechanisms

As the core element of information cocoons, users constitute the main focus of impact mechanism research. Since Harvard professor Cass Sunstein first introduced the concept in Infotopia: How Many Minds Produce Knowledge, the majority of scholars have focused their impact research on the negative effects of information cocoons based on users' selective exposure psychology. According to the scope of research subjects, this research can be summarized into two aspects.

First, research on information cocoon impact mechanisms targeting broad user populations. Scholars have approached from the perspective of users' right to know, indicating that information cocoons pose serious negative impacts on this right. Jiang Xiaoling et al. analyzed from the user reading perspective how algorithmic recommendation technology narrows users' reading horizons, creating a so-called information cocoon that further forcibly绑架 users, creating a vicious cycle. Zhou Chuanhu constructed an analysis of information cocoons' impact on social consensus cohesion from macro, meso, and micro levels and proposed relevant regulatory recommendations.

Second, research targeting college student populations has also become a trend. As a highly active online group with high media usage behavior, college students serve as ideal research subjects that can largely avoid errors caused by non-research variables. Moreover, the development and employment of college students are topics of high social concern, making the exploration of information cocoon impact mechanisms on this group and the mitigation of their side effects socially significant. Wang Huifang argued from the micro-era perspective that information cocoons have very serious negative effects on college students who heavily rely on micro-media for information. Zhang Min et al. analyzed the negative effects of information cocoons on college students in the intelligent era from the perspective of information literacy cultivation.

4. Dilemmas in China's Information Cocoon Research

The above discussion demonstrates that China's journalism and communication academia has developed a considerable body of information cocoon research, with increasingly diversified perspectives as studies deepen. However, objectively speaking, China's information cocoon research still faces two major problems.

First, insufficient development of localized theoretical research. Some scholars have pointed out that "for a long time, a research framework and perspective known as social and behavioral sciences has been set by the West. Of course, this setting of theirs is neither consistent nor accomplished overnight, but has evolved and been refined over thousands of years before gradually being confirmed, disseminated, and diffused, gaining its academic ruling position worldwide." Currently, China's information cocoon research largely continues to follow Western theories. While these research findings have certain significance for promoting the analysis of China's information cocoon generation and impact mechanisms, especially regarding negative effect factors and corresponding countermeasures, Western academia often focuses on individual information selection behavior, emphasizing "binding" personal biases while neglecting the social and cultural complexity of information cocoons. When Chinese scholars cite Western theories, they often fail to fully consider the particularities of local society, potentially resulting in findings that cannot accurately reflect China's information cocoon phenomena. China's traditional culture and media system differ significantly from Western countries, and simply transplanting Western research frameworks into China's information cocoon research may largely lead to the loss of unique local research value.

Second, a late start and singular research perspective. China's information cocoon research began in 2010, while the concept was proposed abroad in 2006. However, some Chinese scholars argue that information cocoons have always existed, noting that Shannon warned the scientific community against overly broad application of information theory to all types of human communication, and Hovland's individual differences theory already elaborated on audience "selectivity and attentionality," meaning information cocoon effects had already appeared and been continuously noted by scholars before the concept emerged. Yet the concept was only introduced to China four years later with the emergence of platforms like Weibo. As discussed in Section 2, China's information cocoon research has consistently focused on aggregated news push platforms and social media, with people-centered research also based on studying users' social media usage behavior. This media technology determinism perspective leads to the neglect of culture and institutions.

Currently, China's journalism and communication discipline's information cocoon research generally revolves around the two cores of "technology" and "people," but overall, it still suffers from the loss of localized theoretical research value under a Western-centric perspective and the problem of singular research perspective under media technology determinism. Although some scholars have touched upon the formation reasons of information cocoons in rural China when studying coffin collection from the perspective of local villagers' information sources, this was only mentioned in passing without in-depth exploration. Thus, we can see that Chinese communication scholars have begun to notice information cocoon effects in interpersonal communication, and we can further expand research on China's local interpersonal communication information cocoon phenomena and generation mechanisms.

5. Possible Paths for China's Information Cocoon Research

China's information cocoon research is still in a developmental stage, requiring long-term exploration regarding theories, methods, and topics, especially in transcending Western research frameworks and expanding research perspectives. Based on the above discussion, this study proposes two possible developmental paths for information cocoon research.

5.1 Breaking Western Academic Shackles

According to visualization results and literature review, China's information cocoon research frameworks and approaches largely remain attached to Western information cocoon research. However, this research system may cause the loss of localized research value in China's context. The negative effects of information cocoons on users significantly outweigh the positive effects, with problems such as group polarization caused by information cocoons adversely affecting social development. Therefore, constructing an information cocoon research system with Chinese characteristics is not only academically significant but also practically important. Some scholars believe that the term "information cocoon" has been conceptualized—in Sunstein's definition, information cocoon theory should be a metaphor, yet most published papers in China use "information cocoon theory" as a title within a Western framework. This phenomenon should prompt reflection in the academic community.

5.2 Transcending the Technological Determinism Research Perspective

Some scholars note that so-called media technology determinism refers to humans being suspended in an environment constructed by media technology, with their concepts and behaviors constrained by the mediated environment and reconstructed due to technological reforms. Information cocoon theory research exhibits this tendency—whenever new media technology emerges, studying the causes and characteristics of information cocoons in that technical context becomes a hot topic for scholars in this field, from algorithm technology and big data to artificial intelligence, repeating in cycles. However, according to Sunstein's definition, information cocoons refer to users in information communication only paying attention to their own choices and pleasant topics, creating a personalized "Daily Me" and thereby rejecting or ignoring other viewpoints and content. Yet information acquisition forms are diverse, and verification and mechanism exploration of information cocoon effects in offline interpersonal communication are lacking, with research perspectives such as culture and institutions facing the threat of being obscured. Therefore, we should break free from the media technology determinism research perspective and explore new research perspectives such as culture and institutions, while constructing a research system for information cocoons that conforms to Chinese characteristics and exploring the theoretical value of this theory in the Chinese context.

References

[1] Yu Guoming, Qu Hui. The Misinterpretation of "Information Cocoon" and the Necessity of Algorithmic Push—On the Solution to Social Ethical Dilemmas in Content Distribution[J]. Journal of Xinjiang Normal University (Philosophy and Social Sciences Edition), 2020(1): 127-133.

[2] Li Zhenhui, Zhu Tingting. Information Diffusion and Content Generation: Research on the Development of Personalized News Platforms[J]. Modern Communication—Journal of Communication University of China, 2017, 39(3): 160-162.

[3] Wang Qian. Opening the "Black Box" of Algorithmic Distribution—A Quantitative Study Based on Toutiao News Push[J]. Shanghai Journalism Review, 2017(9): 7-14.

[4] Wang Yicheng, Wang Ping, Zhang Lu, et al. Research on "Information Cocoon" of Content Intelligent Distribution Platforms from the Perspective of Network Information Ecology Chain[J]. Library Science Research, 2018(9): 7-13, 36.

[5] Tang Lijia, Zhao Zhiqi. Analysis of Douyin's Communication Methods and Problems from a Big Data Perspective[J]. Editing Studies, 2018(6): 52-56.

[6] Shen Nan. Information Cocoon and Information Fairness in the Algorithm Era[J]. Journal of Xi'an Jiaotong University (Social Sciences Edition), 2020(2): 139-144.

[7] Wang Yaqian. "New Pseudo-environment": Research on the Personified Expression of Short Video Bloggers and Its Impact on Users[J]. China Youth Study, 2020(1): 68-75.

[8] Li Longfei, Zhang Guoliang. The Generation Mechanism and Governance Path of "Information Cocoon" Effect in the Algorithm Era—Based on the Perspective of Information Ecology Theory[J]. E-Government, 2022(9): 51-62.

[9] Shi Yingjie, Gui Yong, Huang Ronggui, et al. The Typology, Mechanism, and Impact of Network Media "Cocoon Effect"—A Mediation Analysis Based on the "Chinese College Students' Social Mentality Survey (2020)"[J]. Journalism & Communication Research, 2022(5): 43-59, 126-127.

[10] [US] Cass Sunstein. Infotopia: How Many Minds Produce Knowledge[M]. Translated by Bi Jingyue. Beijing: Law Press, 2008: 18.

[11] Lin Aijun, Liu Yunhong. Algorithmic Bias and Ethical Regulation in Intelligent News Information Distribution[J]. Journalism University, 2020(1): 29-39, 125-126.

[12] Jiang Xiaoling, Ma Jiayi. The "Violence" of Reading: Rethinking Algorithmic Recommendations on News Clients[J]. China Publishing, 2018(24): 43-46.

[13] Zhou Chuanhu. The Dual Effects of "Information Cocoon" on Social Consensus Cohesion[J]. People's Tribune·Academic Frontier, 2019(23): 130-133.

[14] Wang Huifang. The Negative Impact of Information Cocoon on College Students' Information Reading in the Micro-era[J]. Publishing Wide Angle, 2018(2): 69-71.

[15] Zhang Min, Wang Pengjiao, Meng Xiangyu. How Can College Students in the Intelligent Era Break Through "Information Cocoon"?—Based on the Perspective of Information Literacy Cultivation[J]. Modern Educational Technology, 2021(1): 19-25.

[16] Zhai Xuewei. Human Relations, Face, and the Reproduction of Power[M]. Beijing: Peking University Press, 2023: 6.

[17] Hu Yiqing. Rehabilitating Media Technology Determinism: On a New Perspective of Communication Intellectual History[J]. Modern Communication—Journal of Communication University of China, 2017(1): 34-41.

[18] Lin Yufeng. Coffin Collection in Huang Village: A Deep Description of Communication and Cultural Practices in a Southern Village[J]. Chinese Journal of Journalism & Communication, 2024(5): 69-88.