The task of enriching cross-lingual links is to find articles in different languages but representing the same real-world object between multilingual Wikis. In this paper, we propose a novel Multi-Modal Semantic Matching approach, called MMSM, to enrich cross-lingual links for online Wikis. Specifically, MMSM jointly trains two novel end-to-end neural matching models, Entity Description Matching Model and Entity Image Matching Model, which can utilize entity description and images for the cross-lingual entity matching. To the best of our knowledge, it is the first work to utilize multi-modal information to enrich cross-lingual entity links. In the experiments on three datasets $CE{M}_{ZH-EN}^{Easy},$ $CE{M}_{ZH-EN}^{Challenge}$ and $CE{M}_{FR-EN}^{Easy},$ our approach gets the best performance compared with other baseline approaches.

丰富跨语言链接的任务是查找不同语言但代表多语言Wiki之间相同现实世界对象的文章。在本文中，我们提出了一个新的中号ulti-中号odal小号emantic中号atching方法，称为MMSM，以丰富的在线维基跨语言链接。具体地说，MMSM联合训练了两个新颖的端到端神经匹配模型：实体描述匹配模型和实体图像匹配模型，它们可以利用实体​​描述和图像进行跨语言实体匹配。据我们所知，这是利用多模式信息丰富跨语言实体链接的第一项工作。在三个数据集的实验中$CË{\mathrm{中号}}_{žH-Ëñ}^{Ë\mathrm{一种}sÿ}，$ $CË{\mathrm{中号}}_{žH-Ëñ}^{CH\mathrm{一种}升升ËñGË}$ 和 $CË{\mathrm{中号}}_{F\mathrm{[R}-Ëñ}^{Ë\mathrm{一种}sÿ}，$ 与其他基准方法相比，我们的方法可获得最佳性能。