NERsocial: Efficient Named Entity Recognition Dataset Construction for Human-Robot Interaction Utilizing RapidNER

Adapting named entity recognition (NER) methods to new domains poses\nsignificant challenges. We introduce RapidNER, a framework designed for the\nrapid deployment of NER systems through efficient dataset construction.\nRapidNER operates through three key steps: (1) extracting domain-specific\nsub-graphs and triples from a general knowledge graph, (2) collecting and\nleveraging texts from various sources to build the NERsocial dataset, which\nfocuses on entities typical in human-robot interaction, and (3) implementing an\nannotation scheme using Elasticsearch (ES) to enhance efficiency. NERsocial,\nvalidated by human annotators, includes six entity types, 153K tokens, and\n99.4K sentences, demonstrating RapidNER's capability to expedite dataset\ncreation.\n