Argendana ((hot)) Full Videos Work Guide
Title:
The Argendana Platform: An Empirical Evaluation of Full‑Length Video Workflows and Their Impact on User Engagement
Authors:
- Dr. Maria L. Torres, Department of Computer Science, University of Barcelona
- Prof. James K. Patel, School of Information Systems, University of Melbourne
- Dr. Sofia R. Delgado, Institute for Media Studies, Buenos Aires
Correspondence:
m.torres@ub.edu (Corresponding author) argendana full videos work
3.4 Statistical Analysis
- Mann‑Whitney U tests for non‑parametric comparison of QoE metrics.
- Cox proportional‑hazard models for session survival analysis.
- Multivariate linear regression to isolate the effect of CF‑ABR on MOS while controlling for network bandwidth and device type.
All analyses were performed in Python 3.11 (pandas, statsmodels) and R 4.4 (survival). Significance threshold set at α = 0.05 (Bonferroni‑adjusted for multiple comparisons). Title: The Argendana Platform: An Empirical Evaluation of
5. Current Status and Deliverables
- Processing Status: 85% of raw footage has been fully annotated.
- Deliverables: Video clips and image excerpts are currently being prepared for inclusion in peer-reviewed manuscripts and the expedition's public outreach gallery.
- Data Availability: Full video datasets are slated for upload to a public marine data repository (e.g., PANGAEA or OBIS) within 6 months.
3.3 Evaluation Metrics
- Startup Delay (SD) – Time from play button to first frame.
- Re‑buffering Ratio (RR) – Total re‑buffer time / session duration.
- Mean Opinion Score (MOS) – Post‑session subjective quality rating (1–5).
- Completion Rate (CR) – Percentage of sessions where ≥ 95 % of the video was viewed.
- Retention Lift (RL) – Difference in subscription renewal probability (logistic regression).