Problem
Research questions and friction points this paper is trying to address.
programming languages
resource scarcity
code generation
taxonomy
large language models
A Taxonomy of Programming Languages for Code Generation
🤖 AI Summary
This study addresses the severe imbalance in the distribution of programming languages across existing code corpora and the absence of a systematic resource-tiering framework. It proposes the first reproducible, four-tier classification system for programming language resource abundance, based on token-level statistics from seven major code corpora, enabling quantitative assessment of 646 languages. The analysis reveals that only 1.9% of languages—classified as high-resource—account for 74.6% of all code tokens, while the combined share of 71.7% of low-resource languages constitutes less than 1.0%. This tiered framework establishes the first standardized benchmark for data curation and evaluation of multilingual code generation models.
📝 Abstract
The world's 7,000+ languages vary widely in the availability of resources for NLP, motivating efforts to systematically categorize them by their degree of resourcefulness (Joshi et al., 2020). A similar disparity exists among programming languages (PLs); however, no resource-tier taxonomy has been established for code. As large language models (LLMs) grow increasingly capable of generating code, such a taxonomy becomes essential. To fill this gap, we present the first reproducible PL resource classification, grouping 646 languages into four tiers. We show that only 1.9% of languages (Tier 3, High) account for 74.6% of all tokens in seven major corpora, while 71.7% of languages (Tier 0, Scarce) contribute just 1.0%. Statistical analyses of within-tier inequality, dispersion, and distributional skew confirm that this imbalance is both extreme and systematic. Our results provide a principled framework for dataset curation and tier-aware evaluation of multilingual LLMs.
Innovation
Methods, ideas, or system contributions that make the work stand out.
programming language taxonomy
resource-tier classification
code generation
multilingual LLMs
dataset curation
🔎 Similar Papers
No similar papers found.
