CoLa: Chinese Character Decomposition with Compositional Latent Components

• URL: http://arxiv.org/abs/2506.03798v1
• Date: Wed, 04 Jun 2025 10:06:15 GMT
• Title: CoLa: Chinese Character Decomposition with Compositional Latent Components
• Authors: Fan Shi, Haiyang Yu, Bin Li, Xiangyang Xue,
• Abstract summary: Humans can decompose Chinese characters into compositional components and recombine them to recognize unseen characters.<n>We propose a deep latent variable model that learns Compositional Latent components of Chinese characters (CoLa) without relying on human-defined decomposition schemes.
• Score: 63.01848135436417
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Humans can decompose Chinese characters into compositional components and recombine them to recognize unseen characters. This reflects two cognitive principles: Compositionality, the idea that complex concepts are built on simpler parts; and Learning-to-learn, the ability to learn strategies for decomposing and recombining components to form new concepts. These principles provide inductive biases that support efficient generalization. They are critical to Chinese character recognition (CCR) in solving the zero-shot problem, which results from the common long-tail distribution of Chinese character datasets. Existing methods have made substantial progress in modeling compositionality via predefined radical or stroke decomposition. However, they often ignore the learning-to-learn capability, limiting their ability to generalize beyond human-defined schemes. Inspired by these principles, we propose a deep latent variable model that learns Compositional Latent components of Chinese characters (CoLa) without relying on human-defined decomposition schemes. Recognition and matching can be performed by comparing compositional latent components in the latent space, enabling zero-shot character recognition. The experiments illustrate that CoLa outperforms previous methods in both character the radical zero-shot CCR. Visualization indicates that the learned components can reflect the structure of characters in an interpretable way. Moreover, despite being trained on historical documents, CoLa can analyze components of oracle bone characters, highlighting its cross-dataset generalization ability.

Related papers

• The Impact of Visual Information in Chinese Characters: Evaluating Large Models' Ability to Recognize and Utilize Radicals [17.24821720084663]
  We evaluate Large Language Models' and Vision-Language Models' understanding of visual elements in Chinese characters.<n>Our results reveal that models surprisingly exhibit some, but still limited, knowledge of the visual information.<n>We observe consistent improvement in Part-Of-Speech tagging when providing additional information about radicals.
  arXiv  Detail & Related papers  (2024-10-11T17:30:02Z)
• Graph-guided Cross-composition Feature Disentanglement for Compositional Zero-shot Learning [54.08741382593959]
  Disentanglement of visual features of primitives (i.e., attributes and objects) has shown exceptional results in Compositional Zero-shot Learning (CZSL)<n>It is challenging to learn disentangled primitive features that are general across different compositions.<n>We propose the solution of cross-composition feature disentanglement, which takes multiple primitive-sharing compositions as inputs.
  arXiv  Detail & Related papers  (2024-08-19T08:23:09Z)
• Finding structure in logographic writing with library learning [55.63800121311418]
  We develop a computational framework for discovering structure in a writing system. Our framework discovers known linguistic structures in the Chinese writing system. We demonstrate how a library learning approach may help reveal the fundamental computational principles that underlie the creation of structures in human cognition.
  arXiv  Detail & Related papers  (2024-05-11T04:23:53Z)
• Simple Primitives with Feasibility- and Contextuality-Dependence for Open-World Compositional Zero-shot Learning [86.5258816031722]
  The task of Compositional Zero-Shot Learning (CZSL) is to recognize images of novel state-object compositions that are absent during the training stage. Previous methods of learning compositional embedding have shown effectiveness in closed-world CZSL. In Open-World CZSL (OW-CZSL), their performance tends to degrade significantly due to the large cardinality of possible compositions.
  arXiv  Detail & Related papers  (2022-11-05T12:57:06Z)
• Compositional Generalization in Unsupervised Compositional Representation Learning: A Study on Disentanglement and Emergent Language [48.37815764394315]
  We study three unsupervised representation learning algorithms on two datasets that allow directly testing compositional generalization. We find that directly using the bottleneck representation with simple models and few labels may lead to worse generalization than using representations from layers before or after the learned representation itself. Surprisingly, we find that increasing pressure to produce a disentangled representation produces representations with worse generalization, while representations from EL models show strong compositional generalization.
  arXiv  Detail & Related papers  (2022-10-02T10:35:53Z)
• Zero-Shot Chinese Character Recognition with Stroke-Level Decomposition [37.808021793372504]
  We propose a stroke-based method by decomposing each character into a sequence of strokes. We employ a matching-based strategy to transform the predicted stroke sequence to a specific character. The proposed method can be easily generalized to other languages whose characters can be decomposed into strokes.
  arXiv  Detail & Related papers  (2021-06-22T08:49:03Z)
• A causal view of compositional zero-shot recognition [42.63916938252048]
  People easily recognize new visual categories that are new combinations of known components. This compositional generalization capacity is critical for learning in real-world domains like vision and language. Here we describe an approach for compositional generalization that builds on causal ideas.
  arXiv  Detail & Related papers  (2020-06-25T17:51:22Z)
• Compositional Generalization by Learning Analytical Expressions [87.15737632096378]
  A memory-augmented neural model is connected with analytical expressions to achieve compositional generalization. Experiments on the well-known benchmark SCAN demonstrate that our model seizes a great ability of compositional generalization.
  arXiv  Detail & Related papers  (2020-06-18T15:50:57Z)
• Compositionality and Generalization in Emergent Languages [42.68870559695238]
  We study whether the language emerging in deep multi-agent simulations possesses a similar ability to refer to novel primitive combinations. We find no correlation between the degree of compositionality of an emergent language and its ability to generalize. The more compositional a language is, the more easily it will be picked up by new learners.
  arXiv  Detail & Related papers  (2020-04-20T08:30:14Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.