2M-NER: Contrastive Learning for Multilingual and Multimodal NER with Language and Modal Fusion
This addresses the problem of enhancing NER performance across multiple languages and modalities for applications like entity linking and question answering, though it is incremental as it builds on existing multilingual and multimodal NER research.
The authors tackled the lack of a combined multilingual and multimodal dataset for named entity recognition (NER) by constructing a large-scale dataset with four languages and two modalities, and introduced the 2M-NER model using contrastive learning and multimodal fusion, achieving the highest F1 score compared to baselines.
Named entity recognition (NER) is a fundamental task in natural language processing that involves identifying and classifying entities in sentences into pre-defined types. It plays a crucial role in various research fields, including entity linking, question answering, and online product recommendation. Recent studies have shown that incorporating multilingual and multimodal datasets can enhance the effectiveness of NER. This is due to language transfer learning and the presence of shared implicit features across different modalities. However, the lack of a dataset that combines multilingualism and multimodality has hindered research exploring the combination of these two aspects, as multimodality can help NER in multiple languages simultaneously. In this paper, we aim to address a more challenging task: multilingual and multimodal named entity recognition (MMNER), considering its potential value and influence. Specifically, we construct a large-scale MMNER dataset with four languages (English, French, German and Spanish) and two modalities (text and image). To tackle this challenging MMNER task on the dataset, we introduce a new model called 2M-NER, which aligns the text and image representations using contrastive learning and integrates a multimodal collaboration module to effectively depict the interactions between the two modalities. Extensive experimental results demonstrate that our model achieves the highest F1 score in multilingual and multimodal NER tasks compared to some comparative and representative baselines. Additionally, in a challenging analysis, we discovered that sentence-level alignment interferes a lot with NER models, indicating the higher level of difficulty in our dataset.