Research map / Large language models / AI and society

An XAI Social Media Platform for Teaching K-12 Students AI-Driven Profiling, Clustering, and Engagement-Based Recommending

AAAI 2025 · Large language models › AI and society · approach: Classroom tools and student perceptions · 4 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Presents Somekone, a browser-based Instagram-like XAI classroom tool teaching K-12 students data collection, profiling, engagement, and recommendation, with a learning-analytics case study of 209 students clustered by action logs.

Problem

K-12 students often lack understanding of the data-driven mechanisms (data collection, profiling, engagement metrics, recommendation) behind social media platforms, and existing AI education tools rarely connect individual actions to personal and societal impacts.

Approach

An Instagram-like classroom simulator with a paired live analytics dashboard that shows in real time how a student's browsing is logged, profiled, scored for engagement, and used for recommendations, with classroom-wide projector views showing how actions affect everyone's feeds.

Key results

• Latent profile analysis of 209 learners selected a three-cluster solution (Browsers, Engagement enthusiasts, Selective Engagers) with average posterior probability 0.98 and entropy 0.98.
• The tool was tested with roughly 400 children in dozens of schools, with 12 two-hour test sessions involving 209 children.
• Sequence analysis showed the three usage profiles evolved differently over the activity sessions.

Limitations

• High level of abstraction does not model the complexity and scale of real social media platforms, risking misconceptions and insufficient depth for advanced learners.
• Instagram context excludes learning of other AI types and approaches.
• Evaluation of learning outcomes is left to future research; the case example is a learning-analytics demonstration only.

Experimental setup

Datasets: 727-image feed content set labeled by eighth-graders and curated by researchers action logs from 209 learners in twelve Finnish schools

Metrics: average posterior probability entropy BIC

Builds on: latent profile analysis mclust TraMineR

Closest papers by meaning

• A Toolbox for Modelling Engagement with Educational Videos (AAAI 2024 · similarity 0.87)
• Word2Vec4Kids: Interactive Challenges to Introduce Middle School Students to Word Embeddings (AAAI 2025 · similarity 0.85)
• Learning About Algorithm Auditing in Five Steps: Scaffolding How High School Youth Can Systematically and Critically Evaluate Machine Learning Applications (AAAI 2025 · similarity 0.84)
• iLLuMinaTE: An LLM-XAI Framework Leveraging Social Science Explanation Theories Towards Actionable Student Performance Feedback (AAAI 2025 · similarity 0.84)
• AI, Ethics, and Education: The Pioneering Path of Sidekick Academy (AAAI 2024 · similarity 0.84)
• Understanding K-12 Teachers’ Needs for AI Education: A Survey-Based Study (AAAI 2025 · similarity 0.84)