Abstract: Personalized recommendations significantly influence daily life and represent a key shift in information control toward algorithms. This change necessitates new skills for individuals to perceive, understand, and utilize these algorithms effectively, highlighting an urgent need for research on algorithmic literacy within the context of personalized recommendations. This study concentrated on personalized recommendation contexts, developing a knowledge heuristic intervention to enhance algorithmic literacy. A 4-week longitudinal user experiment involving 30 participants was conducted, with statistical comparisons and analyses of changes in algorithmic literacy, as well as the identification of enhancement patterns through cluster analysis. Before and after the experiment, users showed significant improvements in different dimensions of algorithmic literacy, confirming the effectiveness of the knowledge heuristic intervention. Cluster analysis identified three enhancement patterns of algorithmic literacy: gradual improvement with weak foundations pattern, Knowledge-Skill enhancement with weak motivation pattern, and awareness enhancement with strong motivation pattern. This study further analyzed user characteristics associated with each pattern and proposed tailored knowledge heuristic strategies for enhancing algorithmic literacy based on these characteristics.

Key words: Algorithmic literacy, Personalized recommendation, Knowledge heuristic, Heuristic intervention, Enhancement patterns