KAILAS
KAILAS is a Transformer model based on the RoBERTa architecture, specifically adapted for NASA's Science Mission Directorate applications to annotate Unified Astronomy Thesaurus labels.
Downloads 51
Release Time : 4/25/2025
Model Overview
This model, fine-tuned on NASA SMD-related scientific journals and articles, aims to enhance natural language processing technologies (such as information retrieval and intelligent search), particularly suited for text annotation tasks in the field of astronomy.
Model Features
Astronomy Domain Adaptation
Optimized for NASA's Science Mission Directorate applications, particularly suitable for processing astronomy-related texts.
Large-scale Training Data
Trained on 18,000 high-quality astronomy papers, containing approximately 217 million tokens.
Customized Tokenizer
Utilizes a tokenizer optimized for the astronomy domain, improving the ability to handle professional terminology.
Model Capabilities
Astronomical Text Annotation
Unified Astronomy Thesaurus Recognition
Scientific Literature Information Extraction
Use Cases
Academic Research
Automatic Annotation of Astronomical Literature
Automatically annotates Unified Astronomy Thesaurus labels for astronomy papers, facilitating literature classification and retrieval.
Improves the efficiency and accuracy of astronomical literature retrieval
Scientific Knowledge Graph Construction
Extracts key concepts from scientific literature to build knowledge graphs in the field of astronomy.
Promotes the organization and discovery of astronomical knowledge
Information Retrieval
Intelligent Scientific Search Engine
Enhances the understanding of professional terminology in scientific search engines, improving retrieval relevance.
Increases the efficiency of researchers in obtaining relevant literature
Featured Recommended AI Models