Discerning the Political Communication of Short Video Platforms under Algorithmic Recommendations

: As an increasingly important core technology in digitalization, algorithms are not only the "brain" of artificial intelligence, but also a new form of full power. This paper reflects on the epistemological problem of short videos in the context of cybernetics. Through the study of short videos' communication, it points out that short videos have not become a tool for building network communication, openness and communication as expected, but have gradually become centralized, circled and walled, under the control of the algorithm of the big hand behind, forming an "information cocoon". "This paper combines political science, communication science, and the study of the Internet with the study of the Internet. This paper combines the research methods of political science, communication science and sociology to summarize and sort out the representative issues related to short videos, analyze their current problems and development dilemmas, explore the new path of algorithm on the development of political communication of short videos, and provide reference for improving the openness and participatory nature of media programs.


Introduction
According to LatePost, Tiktok has surpassed one billion daily active users, while the four apps that have surpassed one billion daily users worldwide: Facebook, WhatsApp, Instagram and YouTube are now actively expanding the short video track. Current communication studies on algorithms have described the appropriateness of platforms and how they restructure sociality in various fields, and these these authors have extensively analysed the current global current benefits of major internet platforms for the companies and groups behind them, and the relationship between algorithms and politics [1]. Yet the algorithms behind short videos and political communication are becoming increasingly close as short videos become relevant to people's productive lives. In the field of short videos, the issue of political communication is spreading from traditional politically-centred issues to public issues, and the power of political communication subjects is gradually expanding from top to bottom and from the inside out [2]. As short videos become a new tool of production, they increasingly have a powerful 'rationalising power', developing our habits of finding and understanding information and gradually changing our political orientation and ways of thinking. This article focuses on the fact that while algorithms bring convenience to users, they also gradually lead them to give up their right to informed choice, leading to such prominent problems as homogenisation of information, desensitisation of content and homogenisation of values, and the gradual formation of small groups and circles with clear preferences for information consumption. However, there are not many studies today that explore algorithms, platforms, data, and political communication from the perspective of short videos. The contribution of this paper to this scholarship is to show that these Internet short-video companies are more than just the platform attributes described by these previous scholars; increasingly, the platforms behind short-video have become to include everyday sources of information, educational learning terminals, social platforms, etc. They have become key everyday foundations with scale and universality, and the attitude of the platforms and governments behind them towards algorithms is something to think about At the same time, the secondary problems caused by the "information cocoon" created on the information highway are also worth discussing.

The Algorithmic Recommendation Model of Short-form Video Distribution
Generally speaking, short video platforms do not produce content themselves but rather present and deliver original or reproduced content to a broad audience through a centralized pushing process of selection, distribution, and streaming. Among the vast amount of information, political content and issues are filtered by algorithmic mechanisms to reach the audience. Through interactive reproduction and the rise of likes and views, they are transformed into a traffic pool that attracts public attention and becomes a public issue that triggers widespread discussion. Its dissemination generally goes through four stages: production, dissemination, pushing flow, and dissipation. Each location has its own rules and characteristics [3].

Production of Content Topics
The production of short videos on platforms recommended by algorithms is generally not created natively by the platform. Still, it is selected after being uploaded by users and screened by the platform. Although the forum is open to all users, allowing more content producers to participate in the entire production chain, the algorithm is more like the central nerve of the "brain" role, which connects the mass of information and users. When the algorithms filter the content to meet users' needs and generate hot topics, popular content is created in the two. In the process of developing public political issues, due to the low barrier to entry and openness of the internet, coupled with the decentralization and downward shift of discourse under technological empowerment, all internet users are theoretically qualified to participate in public issues and express their positions, allowing short video platforms to provide instant information feedback and multi-party interaction [4]. Algorithms precisely match users' needs before the information is disseminated and can be adjusted accordingly according to user feedback in all aspects of dissemination. Using big data support, algorithms mine and analyze user information and behavioral data, extract users' political preference characteristics, deeply describe user profiles, precisely match information and user needs, grasp landing points, and strengthen the personalization and customization of their push messages, thus influencing the triggering and generation of popular controversial Content triggering and generation [5].

The Dissemination of Content Topics
Based on the user portrait and accurate matching, the algorithm will recommend to the user the content chosen by other audiences with high similarity. The user's media usage and historical behavior will have a large amount of behavioral data left by the audience on the platform. Through the algorithm's calculations, progressively distributed content allows for the sharing and transfer of experience between similar subjects. However, the algorithmic filtering system also implicitly excludes some topics, depriving users of their freedom of choice in information. This kind of recommendation is similar to that of WeChat's moments, but the difference is that WeChat is mainly based on the solid relationship connection of a process of acquaintances, where the users themselves actively establish the relationship [6]. Whereas the information sharing under algorithmic recommendation is mainly based on the weak relationship connections of strangers, which are randomly selected by the algorithm after collaborative filtering. The interpersonal relationship is continuously strengthened by the algorithm, forming a deeper "information cocoon" [7].

The Sublimation of Content Topics
The mass dissemination of short videos is an essential process for content topics to gain universal meaning and absorb public discussion. At the stage of sublimation and discussion of video content, the attributes of the case are not fixed on top of the original information. Still, they will lead to two possible "variations" of "sublimation" or "distortion." The process of "mutation" is similar to that of "distortion." The method of "mutation" is closely related to public participation and the guidance of the algorithmic platform [8]. If the public generates positive feedback on the video content, the venue combined with the algorithm will guide the topic in a practical and reasonable direction, which will often promote the "sublimation" of the topic attributes; conversely, if negative feedback comments or discussions are generated, the topic will be "sublimated." Conversely, if negative feedback, comments, or discussions are generated, the issue will "lose focus" and deviate from its original attributes. The self-purification mechanism built by information filtering technology and the environmental purification mechanism built by user choice preferences can return public attention and participation to the issue attributes themselves, bringing back rational knowledge [9]. Under the competition mechanism of "traffic owners" on the platform, the algorithm blocks some videos. The video symbols that match the audience's preference are collected in the background. The algorithm adjusts and eliminates the less controversial videos with low clicks and reads so that they cannot enter the next round of the video broadcasting cycle. Audiences can freely access the short video platform to voice and express the video content through actions, such as liking and commenting, and participate in the discussion and dissemination of the video. When the public download and repost the video to other social media or even attach their understanding and comments, the field of public interaction is no longer limited to the area of the short video platform. This secondary dissemination and diffusion often involve exchanging more subjects, bringing together collective wisdom, and realizing the sublimation of the properties of quick video content. On the other hand, the anonymity mechanism and the mentality of "no one is to blame" may also lead to the harmful dissemination of short-form video content. In cyberspace, emotions are more likely to go viral than facts, and extremist, eye-catching comments flood the internet, turning cyberspace into a "carnival" of voices [10].

Content Topics that Go out of Fashion
Popular political short videos after a period of dissemination, some popular short videos will temporarily disappear or permanently out of the public view and platform dissemination range. From the video generation to the departure of the dissipation, there exist two ways inside the audience to actively block the video and the platform to block the dissemination. When short videos are overpushed by algorithms, audiences are generally overexposed to redundant information for a period of time, resulting in information fatigue. In contrast, homogeneous videos cannot sustain public attention. On the other hand, the profitability of commercial platforms blocks the dissemination of videos, especially some sensitive short videos, which are deliberately avoided by algorithmic platforms. This has caused political videos to recede from the platform without the attention and discussion they deserve [11].

Data Colonialism
In everyday life, users are progressively documented by data, and most short-form video apps are dominated by algorithms or data that implicitly operate in the background, invisibly "socially sorted," [12] which more or less carry important value choices that are implicitly or explicitly made by those who construct and implement them. Over time, these values enter our brains and become our values and consciousness. Algorithms were once the underlying technological tools that assisted in the distribution of news, helping us to be able to access information more accurately. In the constant evolution of technology, which is more like the new continents discovered in the opening of the New Road, algorithmic communication has gradually evolved into a political and expansive act, with the once territorial claims slowly transforming into data claims, algorithmic claims, and at the same time a new means of control; the famous British sociologist and media researcher Nick Nick Couldry, a leading British sociologist and media researcher, argues that current global data production practices herald a new phase in the development of colonialism and that what is happening now is not just a continuation of capitalism, but has a more profound logic -it is a new way of distributing the world's resources, a way of promoting economic growth. It is a new way of allocating the world's resources, a new way of accessing resources for economic growth, which is the everyday life of human beings itself. By capturing human experience and transforming it into data with potential economic value, we can extract value from the stream of human experience, a new form of the loot we call data colonialism.

The Narrowing of Information Channels Produces Polarised Feedback
Political short videos are different from other short videos in that they are sensitive and ideologically charged. With algorithmic recommendations, the constant pushing of values, and mechanized commercial pushing, the audience's thinking and judgments are constantly solidified. At the same time, deep inside the "information cocoon," the audience will subconsciously form a "view of reality" that is highly consistent with the values of the video output. In today's globalized and collaborative division of labor, the information cocoon may lead to a reverse globalization phenomenon, where platforms and algorithms make the audience's mindset more and more solidified, turning them into "information islands" and one-way people, invariably setting up walls of isolation, and thus moving from globalization to reverse globalization [13].

Thical Establishment and Norm Building
People create technology, but of course, in its development, it is also embedded with the values and emotions carried behind the developers. as Bhargava, said that Algorithms don't have biases, we do. Nowadays, short-form political communication is an essential tool for shaping political and public perceptions, subtly influencing the direction of public opinion and the public's perception of right and wrong. Some "resource power players" can manipulate individual interests and control the rhythm of public opinion. This invisible manipulation puts the user's spirit at risk of being colonized. Therefore, the paradigm of algorithmically produced short-form video content gate-keeping needs to be structurally integrated, as was the case when the WTO was established, to break down barriers, enhance human subjectivity in the digital information age, and create a common domain between individuals. At the same time, a public goods production model should be developed in order to explore the development path of human-media co-production. Secondly, personalized information production and dissemination should combine ethical values with artificial intelligence. The establishment of ethics should be expanded from journalists to various participants, including users, programming technicians, algorithm engineers, algorithm platform managers. At the same time, it is necessary to establish algorithm ethics to ensure that platforms and users are not blinded and manipulated by algorithms [14].

Towards a New View of the Short Videos
"Every technical architecture, every line of code, every interface represents a choice, implies a judgment, and carries a value." [15] Both creators and short video platforms should shoulder their due responsibility for value. Improving the accuracy of the algorithm audit, but also play the plurality of algorithmic recommendations, to improve the dynamics of audience participation. In terms of content distribution, data and blockchain technology should be used to constantly optimize the recommendation algorithm, and promote the transformation of "personalized" recommendations into "humanized" guidance. This shift can recommend users' favorite political short videos, and pushing out in-depth analysis and diversified commentary at the same time. Based on constantly correcting the single value bias, more "humanized" recommendations will be used to find a balance, allowing audiences to access richer information and form more diverse values [16].

Conclusion
Are algorithms, platforms and data good or bad for social impact? There are a thousand Hamlets for a thousand readers. But one thing is clear: science and technology are a double-edged sword. Algorithms have revolutionised the mode of political communication and feedback, bringing with them the benefits of efficiency and speed, but also a host of problems and endless worries. This notion of media has multiple implications: firstly, history shows that technological progress has changed people's lives in a profound way, and that the establishment of algorithmic ethics and norms, just as the establishment of the Magna Carta laid the cornerstones of democracy and civilisation, is a necessary path for the progress of human society. Secondly, the technological iteration of the media should at the same time pay more attention to the evolution and development of people. Reducing the design of media algorithms in terms of automated decision-making and increasing the openness and accessibility of media processes is also a step in the process of algorithmic democracy. In the case of political communication and algorithms for short videos, the important thing is not to reject it, but to face it, accept it and deal with it.