让知识图谱嵌入模型能预测数值属性，提升真实世界知识表示能力。
Neural Regression with Embeddings for Numerical Attribute Prediction in Knowledge Graphs
- 用神经回归模型直接预测知识图谱中的数值属性
- 在多个数据集上达到最优或第二好性能，尤其在双线性模型上提升显著
- 联合训练框架让模型同时优化链接预测与属性预测，增强语义理解
近年来，归纳式知识图谱嵌入模型被用于链接预测和查询回答等任务。尽管知识图谱包含丰富的数值属性，但多数嵌入模型忽略这些信息，限制了对现实世界知识图谱的表达能力。本文提出一种神经回归模型（LitEm），使归纳式知识图谱嵌入模型能够预测图中数值属性。实验表明，LitEm在FB15K-237、YAGO15K、DB15K和Mutagenesis数据集上的多数属性上达到最佳或第二佳表现。此外，我们提出一种联合训练框架，将先进的归纳式知识图谱嵌入模型与LitEm协同训练，显著提升双线性模型的链接预测性能，同时实现数值属性预测。字面感知评估显示，联合训练使模型以‘字面感知’方式编码和利用属性信息，证明性能提升非单纯由参数增加导致。代码已开源：https://github.com/dice-group/dice-embeddings。
原文摘要 · Abstract (English)
In recent years, transductive knowledge graph embedding models have been applied to tasks such as link prediction and query answering. Although knowledge graphs often contain rich numerical attributes, most embedding models neglect them, limiting their ability to represent real-world knowledge graphs with diverse information. In this work, we propose a neural regression model (LitEm) that enables transductive knowledge graph embedding models to predict numerical attributes within knowledge graphs. Experimental results demonstrate that LitEm achieves the best or second-best results on most attributes across FB15K-237, YAGO15K, DB15K, and Mutagenesis. Furthermore, we propose a co-training framework that jointly trains state-of-the-art transductive knowledge graph embedding models with LitEm, which improves link prediction performance mainly for bilinear models and simultaneously enables them to predict numerical attributes. In addition, the literal-awareness evaluation demonstrates that co-training helps models to encode and exploit attribute information in a "literal-aware'' manner, suggesting that the observed gains are not merely due to additional parameters. We publicly release our implementation at https://github.com/dice-group/dice-embeddings.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。