Smart Sequencing of Social Media Microlearning for Higher Education: A Transparent Constraint-Based Algorithm

Social media can extend learning beyond formal higher education settings, but platform-generated content order does not necessarily create a coherent pathway. This limitation is particularly important for social media learning based on brief microlearning objects. This study developed and computationally evaluated a transparent algorithm for content sequencing without using learner profiles, behavioral records, or institutional data. A controlled benchmark of 24 researcher-constructed objects on misinformation and digital verification was organized across seven themes and 42 prerequisite relations. The proposed algorithm restricted eligibility according to prerequisite satisfaction and cognitive stage, then selected among eligible objects using semantic continuity, difficulty smoothness, thematic coverage, and pedagogical content-type diversity. Its output was compared with random, prerequisite-only, semantic-only, and cognitive-only sequencing through six structural metrics, 1,771 weight configurations, and component-ablation tests. The proposed method produced no prerequisite violations, cognitive regressions, or large upward difficulty jumps. Prerequisite-only sequencing eliminated prerequisite violations but yielded a mean cognitive regression rate of .187, while semantic-only sequencing achieved the highest textual continuity but violated 57.1% of prerequisite relations. Rank correlations with the default sequence ranged from .976 to 1.000, and 537 weight configurations reproduced it exactly. The algorithm transparently balanced structural constraints and preference criteria within the benchmark. The findings verify its specified decision rules but do not establish improved human learning.