Abstract: [Purpose/Significance] Based on the widespread application of personalized recommendation algorithms, this article sorts out the formation mechanism of algorithm literacy（AL）of information vulnerable groups in the process of using short video platforms, identifies the influencing factors of AL of information vulnerable groups, and provides a basis for the cultivation of AL of information vulnerable groups. [Design/Methodology] Using grounded theory, this paper encoded interview data in three stages by the Nvivo12, construct a model of the formation mechanism and influencing factors of AL in the interaction between information vulnerable groups and short video content recommendation algorithms,then carried out an exploratory study on the AL of information vulnerable groups. [Findings/Conclusion] The perception of content recommendation algorithms by information vulnerable groups in the process of using short video platforms is mainly triggered by interactions with the algorithm, changes in user interests, replacement of platform used, and comparisons with others. Additionally, the purpose of using short video platforms and the perceived autonomy directly affect their perception of the algorithm. Furthermore, the components of AL are not independent: users' understanding of the internal logic of recommendation algorithms is influenced by their perception of it. Users with a deeper understanding of content recommendation algorithms tend to maintain a critical attitude and adopt proactive algorithm regulation behaviors when using short video platforms. [Originality/Value] The results of this study contribute to deepen the formation mechanism of AL in short video platform interaction, identify relevant influencing factors, and provide suggestions for the cultivation of information vulnerable groups' AL.

Key words: Information vulnerable groups, Algorithmic literacy, Algorithm regulation, Formation mechanism, Influencing factor