সংরক্ষণ করুন:
| প্রধান লেখক: | , , , |
|---|---|
| বিন্যাস: | Recurso digital |
| ভাষা: | ইংরেজি |
| প্রকাশিত: |
Zenodo
2026
|
| বিষয়গুলি: | |
| অনলাইন ব্যবহার করুন: | https://doi.org/10.5281/zenodo.19795545 |
| ট্যাগগুলো: |
ট্যাগ যুক্ত করুন
কোনো ট্যাগ নেই, প্রথমজন হিসাবে ট্যাগ করুন!
|
সূচিপত্রের সারণি:
- <p>The rapid growth of social media platforms has created a strong need for tools that can predict user engagement before content is published. Traditional metrics such as likes, shares, and comments only provide post-publication feedback and fail to capture the subconscious neural processes that influence user behavior.</p> <p>This paper presents a comprehensive survey of twenty-three research studies across four key domains: fMRI-based deep learning models, neuromarketing, emotion recognition from neural signals, and machine learning-based engagement prediction. The study analyzes methodologies, datasets, and findings from each domain to evaluate the feasibility of predicting neural responses to digital content.</p> <p>The findings suggest that modern brain predictive models, such as TRIBE v2, combined with deep learning techniques, enable high-resolution prediction of neural activity associated with attention, emotion, and reward. Additionally, neuromarketing studies demonstrate that neural responses to content occur within milliseconds and strongly correlate with engagement outcomes.</p> <p>Despite significant progress, a major research gap exists in integrating neural prediction, interpretation, and real-world content evaluation into a unified system. This paper highlights this gap and proposes future directions, including applications in social media optimization, real estate, e-commerce, and ethical frameworks for neural data usage.</p> <p>Overall, the study concludes that neural engagement prediction is no longer a theoretical concept but an achievable system, shifting content evaluation from reactive analytics to proactive prediction.</p>