Core Viewpoint - The conference "Diverse Coexistence and Collaborative Future" highlighted the transformation of the publishing industry through digital technology, emphasizing the need for a complete industry chain from e-reading to publishing and IP derivatives [1][2]. Digital Reading Market Growth - In 2024, China's digital reading user base is projected to reach 670 million, a year-on-year increase of 17.52% - The total number of digital reading works is approximately 63.07 million, growing by 6.28% - The digital reading market revenue is expected to reach 66.141 billion yuan, an increase of 16.65% [2]. Industry Challenges - The Chinese publishing industry faces dual pressures of technological and market transformation, with a critical need to connect quality content with internet channels and vast user bases [2]. - The association has facilitated connections between Tomato Novel and over 400 publishing institutions, with 380,000 e-books available [2]. User Engagement and Preferences - Over 10 million users read e-publications daily on the Tomato Novel platform, with a 70% year-on-year increase in daily active users consuming e-publications in 2024 [3]. - The top five categories of publications read include quality novels, literary classics, personal growth, inspirational content, and Chinese history, indicating a strong demand for engaging stories and cultural enrichment [3]. - Users from lower-tier cities are becoming significant contributors to digital reading consumption, with the highest growth rates observed in these areas [3]. Strategic Directions for Tomato Novel - The company aims to scale up the publication of original online literature in physical formats, focusing on popular genres like youth literature and realistic themes [4]. - It plans to enhance collaboration between authors and publishers to explore diverse promotional strategies for new releases [4]. - Tomato Novel will continue to introduce digital content and expand the audio book market, catering to various user scenarios [4]. Cultural Diversity and User Behavior - The rise of digital reading platforms has democratized access to cultural resources, with free platforms playing a crucial role in this shift [5]. - Readers are increasingly favoring books with coherent structures and logical flow, reflecting a trend towards "shallow" and "fragmented" reading habits [5]. - The popularity of IP-driven content is growing, particularly among younger audiences, indicating a shift in industry focus [5]. Addressing Digital Divide and Information Isolation - Recommendations include collaborative efforts between academia and industry to tackle the digital divide and information isolation issues [6]. - There is a need to lower the operational barriers for elderly users, who represent a rapidly growing demographic in the reading market [6]. - To prevent the formation of information silos, it is suggested that algorithms incorporate cultural and interest diversity to enhance user experience [6].

Core Insights - Algorithm recommendation technology has deeply integrated into various fields and scenarios of economic and social development, enhancing user experience by filtering redundant information and improving information acquisition efficiency [1] - While algorithm recommendations drive user engagement and growth for platforms, they also pose risks such as content homogenization and the emergence of "information cocoons" [1] - The proliferation of low-quality content threatens the content creation ecosystem, potentially leading to the loss of quality creators and users, which could severely impact the healthy development of platforms [1]

Core Viewpoint - The article discusses the challenges and implications of algorithmic recommendation systems, highlighting their dual nature as both beneficial and potentially harmful to users and content creators [1][2][3]. Group 1: Algorithmic Recommendation Systems - Algorithmic recommendation technology has deeply integrated into various sectors, enhancing user experience by filtering redundant information and improving information retrieval efficiency [1]. - However, the misuse of algorithmic recommendations can lead to issues such as content homogeneity, creating "information cocoons" that limit users' exposure to diverse viewpoints [1][2]. Group 2: Impact on Content Creation - The prevalence of low-quality, sensational content driven by a "traffic-first" approach undermines the content creation ecosystem, potentially harming users' critical thinking and values, especially among minors [2]. - The distortion of content creation values can lead to a decline in high-quality content production, as creators may feel pressured to conform to low-quality standards to gain visibility [2]. Group 3: Governance and Future Directions - Current governance efforts have made progress, with platforms introducing features like "cocoon assessment" and "one-click breaking cocoon" to enhance user autonomy and diversify recommended content [3]. - The article emphasizes that algorithm governance is a long-term, systematic endeavor requiring collaboration among various stakeholders to ensure algorithms promote positive content and protect user interests [3].

大概6年前的这个时候，我写了一篇文章《信息茧房是个伪概念》，对当时大热的"信息茧房"概念进行 了质疑。 信息茧房，说的是推荐算法让每个人只能看到自己感兴趣的内容，久而久之，就容易陷入到认知茧房的 包裹中，由此造成共识割裂、群体极化等一系列后果。 在文章里，我指出这个概念缺少实证依据支撑，逻辑层面也立不住脚。最后得出了结论，这是个伪概 念，算法的负面影响被夸大了。 不用看原文，因为我现在想法已经改变了。 这几年，我已经慢慢意识到，算法对这个世界的影响已经深入骨髓，它造成的割裂，不仅仅是信息消费 层面的，它让整个社会都在陷入茧房化。而信息茧房，只是整个社会茧房化一个小小隐喻。 社会的茧房与孤岛化、乃至共识的分裂、群体的分道扬镳，这是互有逻辑关联的一系列问题，背后有相 当复杂的成因，但算法是最直接的推手。 因为算法的存在，所有人的信息消费，从过往阅读同一份报纸，同看一档电视节目，变成各自拿着手 机，接收算法个性化推送的内容。这迫使信息接收从带有一定公共性的活动，退化成一种极为私人化的 体验。 1985年，《新闻联播》的最高收视率可以达到72%；1998年《还珠格格》播出期间，全国犯罪率显著下 降；1983年的春晚， ...

Group 1 - Major companies are increasingly focusing on knowledge base functionalities, particularly in the context of AI advancements and the need for efficient information management [2][3][4] - The knowledge base addresses the urgent need for information transformation in small and medium enterprises, allowing them to systematically store and manage scattered data as digital assets [5][7] - The demand for internal knowledge digitization has surged in the AI era, as companies seek to repurpose previously dormant unstructured data into valuable resources [8][12] Group 2 - While the value of knowledge bases is clear, there are concerns about potential pitfalls, such as the risk of content overload and the creation of information silos within organizations [9][10][31] - Companies may become overly reliant on historical data, which could hinder innovation and responsiveness to market changes, as past data may not accurately predict future trends [12][13] - The management of knowledge bases remains a critical challenge, as maintaining content quality and relevance requires significant human resources [16][19][20] Group 3 - The personalization of knowledge base content raises concerns about creating invisible data divides within organizations, potentially leading to misalignment in cross-departmental collaboration [23][31] - Different departments may interpret data differently based on their unique perspectives, which can complicate decision-making processes and hinder effective teamwork [27][30] - The integration of AI in knowledge management is still in its early stages, with many platforms lacking advanced governance capabilities to ensure content accuracy and relevance [21][22]

Core Viewpoint - The rapid development of information technology and algorithmic recommendations has led to the phenomenon of "information cocoon," where users are limited to homogeneous information, restricting their perspectives and solidifying their cognition [1][3]. Group 1: Algorithmic Recommendations - Algorithms serve as personalized "assistants," providing users with tailored content based on their preferences, which enhances user experience [2][4]. - However, there is a risk of "abuse" of algorithms, leading to a narrow view of the world as users are repeatedly exposed to similar content [4][6]. Group 2: Impact on Different Demographics - Different demographic groups experience the effects of the "information cocoon" differently, with older adults being more susceptible due to their limited familiarity with internet operations and information filtering capabilities [7]. Group 3: Implications for Social Discourse - The existence of "information cocoons" is detrimental to the healthy development of online public opinion, as it leads to a narrow perspective and reinforces biases and stereotypes, potentially escalating the spread of rumors and extreme behaviors [8]. Group 4: Governance and Solutions - The Central Cyberspace Administration of China has been actively enhancing governance of information recommendation algorithms, addressing issues such as the promotion of low-quality content and the exacerbation of "information cocoons" [10]. - Recommendations for overcoming "information cocoons" include improving regulatory frameworks, promoting algorithm transparency, and empowering users with tools to manage their content preferences [10][11].

Group 1 - The central government has launched a two-month special action called "Clear and Optimized Business Network Environment - Rectification of Enterprise-related 'Black Mouths'" to address issues affecting businesses [1] - The action focuses on four major problems: malicious defamation of enterprises, extortion, malicious marketing, and infringement of privacy [2] - Specific issues include the organization of "internet water army" for negative campaigns, extortion through negative information, and the distortion of corporate information for malicious marketing [2] Group 2 - The "Clear and Governance of Typical Algorithm Problems on Network Platforms" initiative aims to address concerns regarding algorithmic recommendations that promote low-quality content and create "information cocoons" [3] - Key platforms have responded by signing the "Algorithm for Good" declaration and are working on optimizing their recommendation algorithms to enhance content diversity [3] - Innovations such as "cocoon assessment" and "one-click breaking cocoon" features are being developed to improve user interest management and content variety [3]

Core Viewpoint - The article discusses the "Clear and Bright" initiative by China's Cyberspace Administration aimed at addressing algorithm-related issues such as the promotion of vulgar content, the creation of "information cocoons," and the polarization of viewpoints through algorithmic recommendations [1][2]. Group 1: Algorithm Governance - The initiative emphasizes the need for major platforms like Douyin and Xiaohongshu to optimize their recommendation algorithms to promote positive content, ensure user choice, enhance content diversity, and improve algorithm transparency [1][2]. - The goal is to transform algorithms from amplifiers of bias into tools for efficiency, thereby reducing user anxiety and fostering a healthier content platform ecosystem [1][3]. Group 2: Misunderstanding of Algorithms - There is a prevalent misunderstanding regarding algorithms, with many believing they only serve to reinforce users' existing preferences and viewpoints, leading to increased polarization and division [2][3]. - The article highlights that the creation of "information cocoons" is not solely a flaw of the algorithms but rather a complex issue influenced by various factors, including user behavior and data representation [2]. Group 3: User Participation and Collaboration - The resolution of "information cocoons" requires not only technological and policy interventions but also active user participation and collaborative efforts from society as a whole [3]. - The article suggests that rather than blaming algorithms for biases, there should be a focus on how to better utilize algorithms to meet human needs [3][4]. Group 4: Embracing Technology - The current situation calls for an open mindset towards technology while maintaining a rigorous and professional approach to its application, enabling innovation and transformation that benefits everyone [4].

文/张传文 超级平台用户规模扩大，商业属性不断强化，同时作为社会基础设施的角色和功能更加明显，算法治理 正在面临新的挑战。有一些专家甚至断言，未来平台将在其内部发挥着"准立法权"、"准司法权"和"准 行政权"，政府、研究机构或媒体难以全面掌握平台企业的用户状况和实际运行情况，其相关行动需要 平台企业给予积极配合才可能完成，实现"穿透式监管"或实质性监督方面存在现实困难。 信息茧房、观点极化……如何解决"困在算法里"，做到"算法向善"？ "清朗·网络平台算法典型问题治理"专项行动开展以来，针对网民反映强烈的算法推荐加热低俗信息、 加剧"信息茧房"、加重观点极化等问题风险，中央网信办督促指导重点平台对信息推荐算法功能进行持 续性和针对性优化，取得阶段性成果，如公开算法规则原理，开发"茧房评估""一键破茧"等功能。但算 法治理仍是一项长期性、系统性工程，需常态化巡查监督和持续优化，中央网信办正在努力探索多样 化"破茧"路径，做到切实维护网民合法权益。 技术在创新，监管创新步伐亦未停止。境内互联网信息服务算法备案信息公开发布，是监管部门在算法 治理领域迈出的重要一步，目前国家互联网信息办公室的公布数量已达数百条。平台 ...

Core Viewpoint - The article discusses the efforts of major online platforms in China to optimize their algorithm recommendation systems in response to government initiatives aimed at addressing issues such as the promotion of vulgar content, the exacerbation of "information cocoons," and the polarization of viewpoints [1][2]. Algorithm Transparency - Major platforms have taken steps to enhance algorithm transparency by publicly disclosing their operational rules and promoting user awareness [3]. - Douyin has established a "Safety and Trust Center" to explain its recommendation logic and governance outcomes [4]. - Weibo has improved the transparency of its trending search algorithms by publicizing ranking rules and data metrics [4]. - WeChat Video Account has utilized simple graphics and videos to inform users about its algorithm recommendations [4]. Breaking "Information Cocoons" - Platforms have introduced innovative features like "Cocoon Assessment" and "One-Click Break Cocoon" to help users mitigate the risks associated with "information cocoons" [5][6]. - Douyin has upgraded its "Usage Management Assistant" to visually present users' recent browsing content [6]. - Xiaohongshu has implemented features for content preference assessment and exploration of diverse recommendations [6]. - Kuaishou has enhanced its algorithm to promote positive and useful content [6]. Content Review Mechanism - Platforms are continuously improving their content review mechanisms to strengthen the promotion of positive content and prevent the recommendation of vulgar or harmful information [7][8]. - WeChat Video Account has refined its dual mechanism of "Friend Recommendations" and "Algorithm Recommendations" to better identify and eliminate inappropriate content [8]. - Douyin has introduced a verification mechanism for trending topics to prevent the spread of misleading or fabricated content [8]. User Empowerment - Platforms are optimizing features related to user interest management and content feedback to allow users to adjust their algorithmic recommendations [9]. - Kuaishou offers detailed interest preference management tools for users to customize content delivery [10]. - Weibo has provided various negative feedback options for users to express disinterest in certain content or creators [10]. - The government acknowledges that while progress has been made in algorithm governance, there are still gaps in functionality and content quality that need to be addressed [10].