From Synthetic to Native: Benchmarking Multilingual Intent Classification in Logistics Customer Service

ArXi:2603.23172v1 Announce Type: new Multilingual intent classification is central to customer-service systems on global logistics platforms, where models must process noisy user queries across languages and hierarchical label spaces. Yet most existing multilingual benchmarks rely on machine-translated text, which is typically cleaner and standardized than native customer requests and can. therefore. overestimate real-world robustness. We present a public benchmark for hierarchical multilingual intent classification constructed from real logistics customer-service logs.