CMNER:基于微博的中文多模态实体识别数据集OACSTPCD

Multimodal Named Entity Recognition(MNER)is a pivotal task designed to extract and classify named entities from text with the assistance of pertinent images.Nonetheless,a notable paucity of manual annotation data for Chinese MNER has considerably impeded the progress of Chinese multimodal named entity recognition.We compile a Chinese Multimodal NER dataset(CMNER)utilizing data sourced from social media platform,encompassing 5 000 Weibo posts paired with 18 326 corresponding images.The entities are classified into four distinct categories:person,location,organization,and miscellaneous.We applied the ACN model and UMT model as baseline experiments on CMNER.The experimental results indicate that the F1 scores of the two models reach 74.22%and 89.50%,respectively,validating the effectiveness of the dataset.Furthermore,we conducted cross-lingual experiments and the results substantiate that Chinese and English multimodal NER data can mutually enhance the performance of the NER model.To promote related research on Chinese MNER,the CMNER and related code are released.