A Bolu
收藏资源简介:
A Bolu是由比萨大学和那不勒斯东方大学联合创建的首个撒丁岛即兴诗歌结构化语料库,专注于Logudorese方言的cantada logudoresa诗歌形式。该数据集包含2,835个诗节,共计141,321个词汇标记,数据源自撒丁文化网站làcanas.it的数字化转录文本。语料采用JSON层级结构存储,包含诗人标识、韵律类型、执行时间戳等丰富元数据,并通过去重、实体解析和缺失标注等严格流程进行标准化处理。该资源为计算语言学提供了研究少数语言即兴创作中公式化模式(如Parry-Lord理论)的基准,旨在推动对低资源口头传统的定量分析和NLP工具开发。
A Bolu is the first structured corpus of Sardinian improvised poetry, jointly developed by the University of Pisa and the University of Naples L'Orientale, focusing on the cantada logudoresa poetic form of the Logudorese dialect. This dataset contains 2,835 stanzas, totaling 141,321 lexical tokens, and is sourced from digitized transcriptions hosted on the Sardinian cultural website làcanas.it. The corpus is stored in a hierarchical JSON structure, with rich metadata including poet identifiers, prosodic types, execution timestamps and other related information, and has been standardized through rigorous workflows such as deduplication, entity resolution and missing annotation handling. This resource serves as a benchmark for computational linguistics research on formulaic patterns in minority-language improvised creation (e.g., Parry-Lord theory), aiming to promote quantitative analysis of low-resource oral traditions and the development of NLP tools.

- 1A Bolu: A Structured Dataset for the Computational Analysis of Sardinian Improvisational Poetry比萨大学; 那不勒斯东方大学 · 2026年



