Hybrid Transformer-CNN
for Policy Evaluation
Analyzing public sentiment on government nutrition programs using advanced AI architecture combining IndoBERT transformers with convolutional neural networks.
Research Abstract
This research introduces a novel Hybrid Transformer-CNN architecture that combines the contextual understanding of IndoBERT transformers with the pattern recognition capabilities of Convolutional Neural Networks to analyze public sentiment on government nutrition programs.
By processing social media discourse and policy documents, our model provides actionable insights for policymakers, enabling data-driven decisions that align with public needs and contribute to achieving UN Sustainable Development Goals 2 (Zero Hunger) and 3 (Good Health and Well-being).
Transformer Power
IndoBERT contextual embeddings capture nuanced sentiment in Indonesian text with state-of-the-art accuracy.
CNN Efficiency
Convolutional layers extract local patterns and features, enhancing classification performance.
Policy Impact
Real-time sentiment analysis enables responsive policy adjustments aligned with public needs.
Methodology Pipeline
A seamless flow from raw data to actionable policy insights
Data Collection
Social media & policy docs
IndoBERT Processing
Contextual embeddings
CNN Feature Extraction
Pattern recognition
Policy Insights
Actionable analysis
UN Sustainable Development Goals
Contributing to global development through AI-driven policy analysis
Zero Hunger
Analyzing nutrition program effectiveness to ensure food security and improved nutrition for all.
Good Health
Supporting evidence-based health policies through sentiment analysis of public health programs.