Empowering Weak Languages Through Cross-Language Hyperlink Recommendation

Wikipedia is an important platform for promoting language inclusivity and sharing global knowledge. However, while languages with more resources have a lot of content, languages with fewer resources face challenges in accessibility and cultural representation. To help address this gap, we use multilingual datasets and neural graph collaborative filtering to recommend missing hyperlinks, helping to improve low-resource languages on Wikipedia. By encouraging cross-language collaboration, this method strengthens the connections and content of these languages, promoting cultural sustainability and digital inclusion. Experimental results show significant improvement in recommendation quality, with clear benefits for weaker languages. This highlights the role of recommender systems in preserving unique cultural aspects, building connections between language communities, and supporting fair knowledge sharing in a globalized world.