Low-Resource Language
A language that lacks sufficient text corpora and computational resources in the digital world. The vast majority of the approximately 7,000 currently existing languages worldwide fall into this category, making low-resource languages one of the core research challenges in NLP and machine translation.
Core Facts
Timeline (last 90 days)
Kikuyu is a low-resource language that mainstream large models usually support poorly, and building a model for it faces the dual difficulty of lacking parallel corpora and evaluation benchmarks.
Existing math reasoning evaluations are almost entirely concentrated on high-resource languages such as English, leaving coverage of low-resource languages—especially Arabic dialects—nearly blank.
DAR-MGSM provides a reusable evaluation paradigm: when evaluating reasoning ability in low-resource languages, language variant and cultural context should be measured as independent dimensions.
Only about 5% of the world's population speaks English at home, but about 63.7% of online text is in English, and models trained on this distribution perform noticeably worse on non-English prompts.
Chilean Spanish belongs to the category of 'Low-Resource Variants'—an underrepresented branch within a high-resource language—which is distinct from a 'low-resource language.'
Low-resource languages refer to languages or dialects that lack large-scale annotated corpora, and publicly available data for many regional dialects is often only tens of hours or even less.
In the NLP field, low-resource languages refer to languages lacking large-scale annotated corpora, pre-trained models, and benchmark datasets, and Nepali falls into this category.
How to ensure model quality for low-resource languages under conditions of data scarcity is a core challenge, while cultural misinterpretation must also be avoided.
Research approaches to the low-resource language problem include speech-first models, community volunteer crowdsourced annotation, and unsupervised cross-lingual representation learning.
Nüshu represents a classic low-resource language challenge for AI, as surviving texts are extremely scarce and a standardized digital corpus is lacking.
1 more timeline events
All Facts (12)
There are approximately 7,000 languages in the world, and the vast majority are low-resource languages.
90%UnverifiedIn the NLP field, low-resource languages refer to languages lacking large-scale annotated corpora, pre-trained models, and benchmark datasets, and Nepali falls into this category.
70%UnverifiedKikuyu is a low-resource language that mainstream large models usually support poorly, and building a model for it faces the dual difficulty of lacking parallel corpora and evaluation benchmarks.
50%UnverifiedDAR-MGSM provides a reusable evaluation paradigm: when evaluating reasoning ability in low-resource languages, language variant and cultural context should be measured as independent dimensions.
50%UnverifiedExisting math reasoning evaluations are almost entirely concentrated on high-resource languages such as English, leaving coverage of low-resource languages—especially Arabic dialects—nearly blank.
50%UnverifiedOnly about 5% of the world's population speaks English at home, but about 63.7% of online text is in English, and models trained on this distribution perform noticeably worse on non-English prompts.
50%UnverifiedChilean Spanish belongs to the category of 'Low-Resource Variants'—an underrepresented branch within a high-resource language—which is distinct from a 'low-resource language.'
50%UnverifiedLow-resource languages refer to languages or dialects that lack large-scale annotated corpora, and publicly available data for many regional dialects is often only tens of hours or even less.
50%UnverifiedResearch approaches to the low-resource language problem include speech-first models, community volunteer crowdsourced annotation, and unsupervised cross-lingual representation learning.
50%UnverifiedHow to ensure model quality for low-resource languages under conditions of data scarcity is a core challenge, while cultural misinterpretation must also be avoided.
50%UnverifiedNüshu represents a classic low-resource language challenge for AI, as surviving texts are extremely scarce and a standardized digital corpus is lacking.
50%UnverifiedAs low-resource language processing technology matures, AI may play an increasingly important role in the rescue of endangered languages and scripts.
50%