Hybrid RoBERTa-GRU Model for Semantic Text Analysis and Pattern Discovery in Healthcare Data Mining

View at publisher Summarise with VIRAT
Authors: Sridevi. S, Chandrasekhar Rohith Bhat
Cited by: 0
References: 15

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

The unstructured textual data on digital platforms has grown exponentially, which has led to an international need to find intelligent systems of data mining, which support the Sustainable Development Goals (SDGs), especially SDG-9 (Industry, Innovation, and Infrastructure) and SDG-16 (Peace, Justice, and Strong Institutions). Conventional data mining methods rely on manually designed features and can hardly reveal rich semantic information within complicated language forms which diminish their performance in the practical world. In order to overcome these drawbacks, this study suggests Hybrid RoBERTa-GRU deep learning model suitable to the needs of efficient semantic text processing and the discovery of relevant patterns. RoBERTa is used to obtain contextual embeddings through dynamic self-attention and the GRU layer optimizes the sequential dependency learning with fewer computational resources. Attention control mechanism is incorporated to place more emphasis on important linguistic items so as to have better decision performance. Multilingual text datasets evaluated using experimental evaluation show a better accuracy of 96.18, high precision, recall and F1-score, compared to standard and single transformer methods. The suggested model will help discover smart knowledge in the areas of cybersecurity, sentiment analytics, decision-making in the government, and responsible digital governance, which helps to sustain data-driven innovation. The research forms a solid basis regarding the future developments of automated text mining systems and data analytics on SDGs.

Fields of Study & Subjects

Topic Modeling Sentiment Analysis and Opinion Mining Mental Health via Writing

Publication details

Journal / source
International Conference on Communication, Computing and Electronics Systems (ICCCES)
Pages
1764–1768
Publication date
21 January 2026
Language
EN