A Bolu: A Structured Dataset for the Computational Analysis of Sardinian Improvisational Poetry

2026-04-21 · Source: Computation and Language · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Expert, quick

Summary

A Bolu is introduced as the first structured corpus specifically designed for the computational analysis of Sardinian extemporaneous poetry, known as cantada logudorese. This dataset addresses a significant gap in Natural Language Processing (NLP) for minority languages and the preservation of oral linguistic heritage. Comprising 2,835 stanzas and 141,321 tokens, A Bolu facilitates the documentation and analysis of improvised poetic structures. The study details the corpus architecture and employs a multidimensional analysis, combining descriptive statistics with computational linguistics techniques. Initial findings reveal recurring patterns in Sardinian extemporaneous poetry, supporting Parry and Lord's theory of formulaicity, and contribute to developing more inclusive NLP tools for less widely spoken languages.

Key takeaway

For NLP Engineers and Research Scientists focused on linguistic diversity, A Bolu offers a critical resource for exploring extemporaneous poetry. Your work can now leverage this structured dataset to develop more inclusive NLP models, particularly for minority languages, and to computationally validate theories of oral creativity like formulaicity. Consider integrating similar corpus creation methodologies for other under-resourced oral traditions.

Key insights

A Bolu is the first structured corpus for analyzing Sardinian extemporaneous poetry, revealing formulaic patterns.

Principles

• Oral heritage needs specific computational resources.
• Improvised poetry exhibits recurring formulaic patterns.

Method

The study uses a multidimensional analysis, combining descriptive statistical indices and computational linguistics techniques to map poetic text characteristics within the A Bolu corpus.

In practice

• Develop NLP tools for minority languages.
• Analyze formulaicity in oral traditions.

Topics

• A Bolu Dataset
• Sardinian Poetry
• Computational Linguistics
• Minority Languages NLP
• Oral Linguistic Heritage

Best for: NLP Engineer, Research Scientist, AI Scientist

Related on AIssential

• AfriSUD: A Dependency Treebank Collection for Evaluating Models on African Languages — Models exhibit a significant "syntax gap" in processing African languages, struggling with their structural diversity.
• Automatic Metrical Scansion of Galician Poetry: First Results — A new system provides automatic metrical scansion for Galician poetry with 88% accuracy.
• Curation of a Palaeohispanic Dataset for Machine Learning — A new structured dataset enables machine learning analysis of under-resourced Palaeohispanic languages.
• Retrato_Cantado: Criação e Análise de um Corpus para Representações de Identidade em Letras de Músicas Brasileiras — Lexical-syntactic patterns effectively identify predicative sentences in song lyrics for identity representation analysis.
• Audience Engagement with Arabic Women's Social Empowerment and Wellbeing: A Decadal Corpus — The corpus provides a decadal view of Arabic women's social empowerment discourse on Facebook.
• Mark Neumann: ScispaCy: A spaCy pipeline & models for scientific & biomedical text — SciSpacy provides a fast, domain-specific NLP pipeline for biomedical text, overcoming general spaCy limitations with specialized models and components.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Computation and Language.