Tag: named entity recognition

A 2026 computational study found that a domain-specific BiGRU-CRF pipeline achieved 96.52% entity-level exact-match accuracy across 6,000 Chinese power-grid text ...

A comparative study of five neural architectures on two datasets found that ELECTRA had the highest full-data accuracy for token-level ...

Researchers at the Research Foundation for SUNY have developed SafeSeal, a patent-pending watermarking system that embeds detectable marks in large ...

Researchers have unveiled Atenea, an open-source platform that continuously archives public Telegram channels, enriches messages with NLP, and enables semantic ...

Researchers at the University of Kashan have developed GGIP2P, a modular pipeline that combines BERT-based entity recognition, pronoun resolution, spatial ...

A new framework called MR-UIE trains large language models to reason from multiple perspectives before extracting structured information, setting new ...

Researchers in China combined a fine-tuned large language model with a TWTA fault knowledge graph to achieve a 94.56 percent ...

Researchers used BERT-based natural language processing on 42,852 LinkedIn comments to reveal that employee discourse shifted sharply toward work-life balance ...