The BALAGHA Score Project

Digital Humanities · Quantitative Linguistics · Arabic Rhetoric

DOI: 10.64393/balagha

The BALAGHA Score Project is a unified research ecosystem for Arabic rhetoric ("al-balāgha"), combining an open encyclopedia, corpus resources, computational tools, a knowledge base, and metrics for rhetorical analysis.


al-Balagha.com

Encyclopedia of Arabic Rhetoric (concepts, devices, examples).


Balagha Corpus

Texts, annotations, and datasets for rhetorical study.

Open source resource:

* Corpus Navigator


The BALAGHA Score

Analytical tools and metrics for rhetorical richness.

Open source resource:

* Arabic Word Tokenisation Scheme


BalaghaBase

Wikibase knowledge graph for devices, scholars, sources.


Rhetorical Density

Methodology and discussion of density as a quantitative metric.


BalaghaID App

Arabic Rhetorical Device Identification with an Interactive App.

Open source resource:

* BalaghaID App

Publications

  1. Mandar Marathe. "The BALAGHA Score: A Quantitative Linguistics Approach to Arabic Rhetorical Analysis.” Paper presented at the 13th International Quantitative Linguistics Conference (QUALICO 2025), Faculty of Arts, Masaryk University, Brno, Czech Republic, June 26–28, 2025.

    * An international conference where the concept of Rhetorical Density, and the BALAGHA Score which is derived from it, were introduced.

  2. Mandar Marathe. "The BALAGHA Score: A Digital Humanities Approach to Assessing Arabic Rhetoric." Paper presented at "Pathways to the Future: A Digital Humanities Conference", SOAS University of London, London, UK, March 7, 2025.

    * A conference organised by SOAS University of London and Keio University, Tokyo, Japan, where the concept of Rhetorical Density, and the BALAGHA Score which is derived from it, were introduced.

  3. Mandar Marathe. “Creation of a Numerical Scoring System to Objectively Measure and Compare the Level of Rhetoric in Arabic Texts: A Feasibility Study, and A Working Prototype.” MA diss., University of Exeter, 2022. https://doi.org/10.48550/arXiv.2507.21106.
    Also available at https://doi.org/10.5281/zenodo.15765533.

Methodological resources

  1. Mandar Marathe. "Corpus Navigator v0.1.0" Zenodo, 31 January 2026. https://doi.org/10.5281/zenodo.18437018.
    Also available on Balagha-Corpus.com.
    Direct download of the Corpus Navigator.

    * The Corpus Navigator is a word position indexer tool for corpus annotations of textual content. This tool implements standardized rules for numbering words in any text, supporting multiple languages and writing systems. It allows text annotators to reliably and consistently identify word positions within texts.

  2. Mandar Marathe. "Arabic Rhetorical Device Identifier v0.1.0" Zenodo, 13 January 2026. https://doi.org/10.5281/zenodo.18236000.
    Also available on GitHub.
    Direct download of the full custom prompt.

    * The Arabic Rhetorical Device Identifier is a custom GPT and AI system prompt package to assist in identifying Arabic rhetorical devices, in classical and contemporary Arabic texts. It is based on the Arabic Rhetorical Device Taxonomy v0.1.1. Please view the full README.md file for usage notes and instructions.

  3. Mandar Marathe. "BalaghaID.app: Arabic Rhetorical Device Identification with an Interactive App v0.1.0" Zenodo, 15 December 2025. https://doi.org/10.5281/zenodo.17935325.
    Also available on GitHub and the BalaghaID website.
    Direct download of the offline version of the BalaghaID app.

    * BalaghaID app is an interactive app for identifying Arabic rhetorical devices through guided decision-making. The system implements a structured decision tree derived from the Arabic Rhetorical Device Taxonomy, guiding users step by step through diagnostic questions that progressively narrow down the rhetorical phenomenon under analysis.

  4. Mandar Marathe. "BALAGHAScore.com Arabic Word Tokenisation Scheme v0.1.0." Zenodo, 10 December 2025. https://doi.org/10.5281/zenodo.17877973.
    Also available on GitHub.
    Direct download of the Tokenisation Scheme PDF.

    * The Arabic Word Tokenisation Scheme is a formal set of rules for consistently segmenting Arabic text into meaningful word units, optimised for the calculation of the BALAGHA Score — a quantitative metric of rhetorical density in both classical and contemporary Arabic texts.

  5. Mandar Marathe. "Arabic Rhetorical Device Taxonomy (v0.1.1) of the Encyclopedia of Arabic Rhetoric (al-Balagha.com) (v0.1.1)." Zenodo, 20 November 2025. https://doi.org/10.5281/zenodo.17663490.
    Also available on GitHub.
    Direct download of the Taxonomy PDF file.

    * Consisting of 95 Arabic rhetorical devices, their definitions, taxonomy, classification, relationships, and usage examples, the Arabic Rhetorical Device Taxonomy is a permanent, shareable, and citable resource which can be used by anyone working in the field of Arabic rhetoric.