Abstract
Now, one of the most important memory institutions, namely social media, has taken on the structuring of collective memory. Social media uses artificial intelligence and machine learning algorithms to analyze data and information recorded from users or big data. Algorithmic selection of social media by classifying, highlighting, deleting and heterogeneity in the level of visibility of users' content, as an application of memory politics that limits and defines narratives for remembering and forgetting. And by creating gaps in memory, it causes collective forgetting and the disintegration of collective memory. The present interdisciplinary research, with a qualitative approach in the tradition of cultural studies, using the grounded theory method and Strauss' coding model, with theoretical and purposeful sampling of interview texts, Instagram, articles and books of collective memory and observation of related activities and events, has examined the algorithmic memory politics and its impact on the postmodern discourse of collective memory in Iran. The findings indicate that the central phenomenon of the research is related to all the main categories of power relations, digital knowledge, ideology and global inequality, memory of resistance, and the crisis of identity and meaning in the postmodern social situation.

Keywords: Algorithmic memory politics, Postmodern collective memory discourse, Cultural studies, Grounded theory method, Artificial intelligence algorithms, Social media









Introduction
According to a report by the reputable website Datareportal (datareportal.com), in 2024, Instagram social media topped the list of the most used social media in Iran with more than 46 million registered users and 31 million active users.
Instagram has become an effective arena for expressing new and personal narratives of the history of Iran and the world and creating new methods of identification and has provided a vast opportunity to revise collective memory. With the increase in the number of users, social media has become more powerful in analysis, decision-making, political polarization and influencing public opinion due to the collection, recording and analysis of user data or big data. Big data is also used to predict future events, unrest and emerging social movements with the help of algorithmic calculations. Automatic social media algorithms select and prioritize data to present to the user, and by displaying or not displaying, classifying, and creating asymmetry and inequality in access to information, they can potentially prevent users from using all the information available on the network. Also, the need to pay attention to the position of interdisciplinary studies in communication sciences, cultural studies, and media has led the present study to examine the politics of algorithmic memory and its impact on the postmodern discourse of collective memory in Iran.

Materials and Methods
This qualitative research was conducted based on the tradition of cultural studies and using the grounded theory method. Data collection was done through a library method and to ensure the validity and reliability of the research, theoretical and purposive sampling was used until theoretical saturation was reached. Coding was done on sixteen interview texts with relevant experts, whose interview texts have been published in research journals as well as on specialized computer science and communication websites. The interview texts were used as secondary data for data collection. In the open coding stage, 130 codes were extracted using the Glasser and Strauss questioning framework. Then, axial coding was done based on Strauss's coding paradigm.
Causal conditions (information capitalism economy, political control of social media, symbolic power of social media and technical and network causes), contextual conditions (diversity and plurality of media, sources and places of memory in the postmodern social situation). The intervening conditions (ideology and global inequality), the central category (algorithmic memory politics), the strategies of action and interaction (memory of resistance and digital struggle), as well as the consequences of action and interaction (conflict of memories, sense of alienation, and crisis of identity and meaning) were separately identified. The politics of algorithmic memory was identified from the selective coding stage as the main phenomenon of the research and by setting the propositions, the main categories were linked to it. Then, the propositions were expanded in the form of a new theory and expressed as the findings of the research.

Discussion & Result
As one of the most important institutions for the structure of collective memory, social media uses automated algorithms, algorithmic calculations, artificial intelligence, and machine learning to select and aggregate items to be remembered or forgotten.
Therefore, a kind of memory policy called algorithmic memory policy is applied to access the content produced and shared by users on social media. In the postmodern discourse of collective memory, collective memory controlled by state memory policy as well as algorithmic memory policy that is constructed to legitimize power relations and under the rule of ideological values ​​and instrumental rationality will be a disjointed, multiple, and incoherent memory whose collective validity is contested and doubted, except in times of collective disasters and traumatic situations. This in itself creates the basis for the diversity and proliferation of memories and ultimately exacerbates memory conflicts and crises of identity and meaning.
This has paved the way for the spread of all kinds of extremism, cultural particularism, and religious fundamentalism in pursuit of ethnic, racial, sexual, etc. identities, and despite the increase in the production of political content among users, it has led to consumerism and apoliticality.

Conclusion
As automated social media algorithms, based on observability based on the attention economy and machine learning, act as memory policies and apply a kind of content selection policy in presenting content to users. Which content is deleted, made invisible, which content is displayed to which user and how many times, it is crucial that we more carefully examine and research the authenticity of the content presented and not make social networks the only authoritative and reliable source as the narrator of our historical past.