Abstract:Chinese Named Entity Recognition (NER) is a very important subtask in the information extraction domain. Its purpose is to locate named entities in the text and classify them into predetermined categories. The key point of NER is to learn a high-quality representation of tokens. Recently, representation learning techniques have been introduced into NER due to their excellent performance in mining the semantics of texts and mastering their organization. In the field of Chinese, many studies introduce multimodal feature extraction schemes to enrich token representations, such as radical and word. However, the learning scheme of these auxiliary features is relatively complex and has difficulty learning the interactive relationship between features with a concatenation or MLP fusion method. To address these challenges, a Multimodal Chinese NER Model based on Self-attention Mechanism named MNER is proposed, which consists of a multimodal feature fusion module and an entity classification module. To study the informative characterization of tokens, a multimodal feature fusion module is proposed to exploit radical, character, and word information. In the multimodal feature fusion module, a self-attention mechanism is designed to integrate multimodal features based on the correlation between the features, which addresses the problem that existing methods have difficulty exploiting interaction information between features. A semantic-aware category modifier is proposed to enhance the CRF classification layer's performance. It increases entity discrimination by adjusting the embeddings of features according to the similarity between the token embeddings and embedding of each entity category, which widens the encoding gap between different entities and narrows the search scope for entity classification. Finally, the proposed MNER is compared with ten state-of-the-art methods on Weibo and Resume datasets, and the results show the superiority of our model on three metrics.

An Improved Method for Chinese Named Entity Recognition Based on MRC

Improving Biomedical Named Entity Recognition with a Unified Multi-Task MRC Framework

A Unified MRC Framework for Named Entity Recognition

Pronounce Differently, Mean Differently: A Multi-Tagging-scheme Learning Method for Chinese NER Integrated with Lexicon and Phonetic Features

Multimodal Features Enhanced Named Entity Recognition Based on Self-Attention Mechanism.

A Multi-task Approach for Machine Reading Comprehension Form Named Entity Recognition Tasks

A Local Information Perception Enhancement–Based Method for Chinese NER

Named Entity Recognition via Machine Reading Comprehension: A Multi-Task Learning Approach

Enhanced Chinese Domain Named Entity Recognition: An Approach with Lexicon Boundary and Frequency Weight Features

Query-Based Named Entity Recognition

MSFM: Multi-view Semantic Feature Fusion Model for Chinese Named Entity Recognition.

Bag of Tricks for Chinese Named Entity Recognition.

Named entity recognition model based on Multi‐BiLSTM and competition mechanism

Named Entity Recognition Based on Pre-training Model and Multi-head Attention Mechanism

Chinese Named Entity Recognition Method Combining ALBERT and a Local Adversarial Training and Adding Attention Mechanism

Enhanced Chinese named entity recognition with multi-granularity BERT adapter and efficient global pointer

A Chinese Named Entity Recognition System with Neural Networks

Pretraining Multi-modal Representations for Chinese NER Task with Cross-Modality Attention

Named entity recognition for Chinese based on global pointer and adversarial training

A Nested Named Entity Recognition Method for Traditional Chinese Medicine Records

MFE-NER: Multi-feature Fusion Embedding for Chinese Named Entity Recognition