{"id":2,"date":"2026-05-03T08:14:03","date_gmt":"2026-05-03T08:14:03","guid":{"rendered":"https:\/\/wp.lancs.ac.uk\/aragenre\/?page_id=2"},"modified":"2026-05-18T03:37:39","modified_gmt":"2026-05-18T03:37:39","slug":"aragenre2026","status":"publish","type":"page","link":"https:\/\/wp.lancs.ac.uk\/aragenre\/","title":{"rendered":"AraGenre 2026"},"content":{"rendered":"

AraGenre: A Hierarchical Definition-Guided Arabic Genre Classification Shared Task<\/strong><\/h2>\n

Mo El-Haj, Saad Ezzini, Mustafa Jarrar, Shadi Abudalfa<\/b>
\n<\/b>m.el-haj@lancaster.ac.uk<\/span><\/a>\u00a0<\/span><\/span><\/p>\n

Registration form: https:\/\/forms.gle\/RAi7SpYoRD5RgUnP6<\/a><\/span><\/h2>\n

AraGenre Codabench: https:\/\/www.codabench.org\/competitions\/16356<\/a><\/span><\/h2>\n

1. Overview<\/h2>\n

\"\"<\/h2>\n

AraGenre is a shared task on hierarchical Arabic genre classification. The task evaluates whether systems can identify the communicative genre of Arabic texts across two levels: Broad Genre<\/strong>, referring to high-level communicative functions such as Informative, Interactive, Creative, Religious, Legal, or Learning, and Specific Genre<\/strong>, referring to finer-grained genre subtypes such as encyclopaedic writing, forum discussion, educational explanation, diplomatic communication, or religious commentary.<\/p>\n

Unlike traditi<\/p>\n

onal text classification benchmarks that rely heavily on large labelled datasets and stable training distributions, AraGenre focuses on robust genre understanding across diverse communicative settings, dialects, orthographic conventions, and writing styles. The task adopts a definition-guided framework where participants are provided with a hierarchical genre taxonomy, English genre definitions, labelled training and development data, and a hidden evaluation benchmark containing naturally occurring Arabic texts.<\/p>\n

The released training and development sets consist primarily of synthetic and carefully controlled examples designed to simulate low-resource conditions. In contrast, the hidden evaluation benchmark contains substantially noisier naturally occurring texts and previously unseen genre-definition combinations, encouraging systems to rely on semantic understanding of communicative function rather than memorisation of narrow lexical patterns.<\/p>\n

The benchmark spans Modern Standard Arabic, Classical Arabic, and multiple Arabic dialects, including both formal and informal writing styles as well as diacritised and undiacritised text.<\/p>\n

Systems are evaluated separately at the broad and specific genre levels. The official ranking metric is Hierarchical Macro F1<\/strong>, calculated as the average of Broad Macro F1 and Specific Macro F1.<\/p>\n

AraGenre is intended as a practical computational benchmarking framework for Arabic NLP. Some categories may overlap with related notions in linguistics and discourse studies, including text type, register, and discourse function.<\/p>\n

Each input instance consists of an Arabic text segment ranging from short fragments to longer passages. Systems must assign both a broad genre label and a fine-grained specific genre label.<\/p>\n

AraGenre adopts a definition-guided evaluation framework where systems receive English genre definitions together with limited labelled training and development data designed to simulate low-resource conditions.<\/p>\n

The hidden evaluation benchmark contains noisier naturally occurring Arabic texts from diverse communicative settings and includes previously unseen genre-definition combinations to encourage semantic genre understanding beyond narrow lexical memorisation.<\/span><\/h2>\n

Table 1: Example Genres with Definitions and Illustrative Samples<\/h2>\n\n\n\n\n\n\n\n\n\n\n
Broad Genre<\/th>\nSpecific Genre<\/th>\nDefinition<\/th>\nExample 1<\/th>\nExample 2<\/th>\n<\/tr>\n<\/thead>\n
Informative<\/td>\nAnalytical Reports<\/td>\nStructured analytical writing discussing trends, developments, policies, or evidence-based observations using factual interpretation or comparative analysis.<\/td>\n\u062a\u0634\u064a\u0631 \u0627\u0644\u0628\u064a\u0627\u0646\u0627\u062a \u0627\u0644\u0627\u0642\u062a\u0635\u0627\u062f\u064a\u0629 \u0627\u0644\u0623\u062e\u064a\u0631\u0629 \u0625\u0644\u0649 \u0627\u0631\u062a\u0641\u0627\u0639 \u0645\u0639\u062f\u0644\u0627\u062a \u0627\u0644\u062a\u0636\u062e\u0645 \u0645\u0642\u0627\u0631\u0646\u0629 \u0628\u0627\u0644\u0639\u0627\u0645 \u0627\u0644\u0633\u0627\u0628\u0642\u060c \u0645\u0645\u0627 \u0623\u062f\u0649 \u0625\u0644\u0649 \u062a\u0631\u0627\u062c\u0639 \u0627\u0644\u0642\u0648\u0629 \u0627\u0644\u0634\u0631\u0627\u0626\u064a\u0629 \u0644\u0644\u0623\u0633\u0631.<\/td>\n\u0623\u0638\u0647\u0631\u062a \u0627\u0644\u062f\u0631\u0627\u0633\u0629 \u0623\u0646 \u0627\u0644\u0627\u0633\u062a\u062e\u062f\u0627\u0645 \u0627\u0644\u0645\u0641\u0631\u0637 \u0644\u0644\u0647\u0648\u0627\u062a\u0641 \u0627\u0644\u0630\u0643\u064a\u0629 \u0628\u064a\u0646 \u0627\u0644\u0645\u0631\u0627\u0647\u0642\u064a\u0646 \u064a\u0631\u062a\u0628\u0637 \u0628\u0627\u0646\u062e\u0641\u0627\u0636 \u0633\u0627\u0639\u0627\u062a \u0627\u0644\u0646\u0648\u0645 \u0648\u0636\u0639\u0641 \u0627\u0644\u062a\u0631\u0643\u064a\u0632 \u0627\u0644\u0623\u0643\u0627\u062f\u064a\u0645\u064a.<\/td>\n<\/tr>\n
Interactive<\/td>\nAdvice Columns<\/td>\nTexts where individuals seek or provide practical, emotional, social, or moral advice regarding personal experiences or everyday situations.<\/td>\n\u0623\u0634\u0639\u0631 \u0628\u0627\u0644\u062a\u0648\u062a\u0631 \u0627\u0644\u0634\u062f\u064a\u062f \u0642\u0628\u0644 \u0627\u0644\u0627\u0645\u062a\u062d\u0627\u0646\u0627\u062a\u060c \u0648\u0644\u0627 \u0623\u0639\u0631\u0641 \u0643\u064a\u0641 \u0623\u062a\u0639\u0627\u0645\u0644 \u0645\u0639 \u0647\u0630\u0627 \u0627\u0644\u0642\u0644\u0642 \u0627\u0644\u0645\u0633\u062a\u0645\u0631\u060c \u0641\u0645\u0627\u0630\u0627 \u062a\u0646\u0635\u062d\u0648\u0646\u0646\u064a\u061f<\/td>\n\u0623\u0639\u0627\u0646\u064a \u0645\u0646 \u0635\u0639\u0648\u0628\u0629 \u0641\u064a \u062a\u0646\u0638\u064a\u0645 \u0648\u0642\u062a\u064a \u0628\u064a\u0646 \u0627\u0644\u0639\u0645\u0644 \u0648\u0627\u0644\u062f\u0631\u0627\u0633\u0629\u060c \u0648\u0623\u062d\u062a\u0627\u062c \u0625\u0644\u0649 \u0646\u0635\u0627\u0626\u062d \u062a\u0633\u0627\u0639\u062f\u0646\u064a \u0639\u0644\u0649 \u062a\u062d\u0642\u064a\u0642 \u0627\u0644\u062a\u0648\u0627\u0632\u0646.<\/td>\n<\/tr>\n
Learning<\/td>\nEducational Explanations<\/td>\nExplanatory educational texts intended to help learners understand academic, scientific, technical, or conceptual topics through clarification and structured reasoning.<\/td>\n\u064a\u062d\u062f\u062b \u0627\u0644\u062a\u0628\u062e\u0631 \u0639\u0646\u062f\u0645\u0627 \u062a\u062a\u062d\u0648\u0644 \u0627\u0644\u0645\u0627\u062f\u0629 \u0645\u0646 \u0627\u0644\u062d\u0627\u0644\u0629 \u0627\u0644\u0633\u0627\u0626\u0644\u0629 \u0625\u0644\u0649 \u0627\u0644\u062d\u0627\u0644\u0629 \u0627\u0644\u063a\u0627\u0632\u064a\u0629 \u0646\u062a\u064a\u062c\u0629 \u0627\u0643\u062a\u0633\u0627\u0628\u0647\u0627 \u0644\u0644\u0637\u0627\u0642\u0629 \u0627\u0644\u062d\u0631\u0627\u0631\u064a\u0629.<\/td>\n\u0644\u062d\u0633\u0627\u0628 \u0645\u0633\u0627\u062d\u0629 \u0627\u0644\u0645\u062b\u0644\u062b\u060c \u0646\u0642\u0648\u0645 \u0628\u0636\u0631\u0628 \u0637\u0648\u0644 \u0627\u0644\u0642\u0627\u0639\u062f\u0629 \u0641\u064a \u0627\u0644\u0627\u0631\u062a\u0641\u0627\u0639 \u062b\u0645 \u0646\u0642\u0633\u0645 \u0627\u0644\u0646\u0627\u062a\u062c \u0639\u0644\u0649 \u0627\u062b\u0646\u064a\u0646.<\/td>\n<\/tr>\n
Legal<\/td>\nDiplomatic Communication<\/td>\nOfficial communication concerning international relations, negotiations, cooperation, agreements, or political coordination between states or organisations.<\/td>\n\u0623\u0643\u062f\u062a \u0627\u0644\u062f\u0648\u0644 \u0627\u0644\u0645\u0634\u0627\u0631\u0643\u0629 \u0623\u0647\u0645\u064a\u0629 \u062a\u0639\u0632\u064a\u0632 \u0627\u0644\u062a\u0639\u0627\u0648\u0646 \u0627\u0644\u0625\u0642\u0644\u064a\u0645\u064a \u0644\u0645\u0648\u0627\u062c\u0647\u0629 \u0627\u0644\u062a\u062d\u062f\u064a\u0627\u062a \u0627\u0644\u0627\u0642\u062a\u0635\u0627\u062f\u064a\u0629 \u0627\u0644\u0645\u0634\u062a\u0631\u0643\u0629.<\/td>\n\u0639\u0642\u062f \u0627\u0644\u0648\u0641\u062f\u0627\u0646 \u0627\u062c\u062a\u0645\u0627\u0639\u0627\u064b \u062b\u0646\u0627\u0626\u064a\u0627\u064b \u0644\u0628\u062d\u062b \u0622\u0644\u064a\u0627\u062a \u062a\u0637\u0648\u064a\u0631 \u0627\u0644\u0639\u0644\u0627\u0642\u0627\u062a \u0627\u0644\u062a\u062c\u0627\u0631\u064a\u0629 \u0628\u064a\u0646 \u0627\u0644\u0628\u0644\u062f\u064a\u0646 \u062e\u0644\u0627\u0644 \u0627\u0644\u0645\u0631\u062d\u0644\u0629 \u0627\u0644\u0645\u0642\u0628\u0644\u0629.<\/td>\n<\/tr>\n
Creative<\/td>\nMotivational Writing<\/td>\nExpressive or reflective writing intended to inspire, encourage, or persuade readers through motivational rhetoric or self-improvement themes.<\/td>\n\u0644\u0627 \u062a\u0633\u0645\u062d \u0644\u0644\u0641\u0634\u0644 \u0623\u0646 \u064a\u0648\u0642\u0641\u0643\u060c \u0641\u0643\u0644 \u062a\u062c\u0631\u0628\u0629 \u0635\u0639\u0628\u0629 \u062a\u0645\u0646\u062d\u0643 \u0641\u0631\u0635\u0629 \u062c\u062f\u064a\u062f\u0629 \u0644\u0644\u0646\u0645\u0648 \u0648\u0627\u0644\u062a\u0639\u0644\u0645.<\/td>\n\u0627\u0644\u0646\u062c\u0627\u062d \u0644\u0627 \u064a\u0623\u062a\u064a \u0635\u062f\u0641\u0629\u060c \u0628\u0644 \u064a\u0628\u062f\u0623 \u0628\u062e\u0637\u0648\u0629 \u0635\u063a\u064a\u0631\u0629 \u0648\u0625\u064a\u0645\u0627\u0646 \u0645\u0633\u062a\u0645\u0631 \u0628\u0642\u062f\u0631\u0627\u062a\u0643.<\/td>\n<\/tr>\n
Religious<\/td>\nReligious Commentary<\/td>\nReflective or explanatory religious writing discussing moral lessons, spiritual guidance, interpretation, or ethical values within a religious context.<\/td>\n\u064a\u062f\u0639\u0648 \u0627\u0644\u0625\u0633\u0644\u0627\u0645 \u0625\u0644\u0649 \u0627\u0644\u0635\u0628\u0631 \u0648\u0627\u0644\u062a\u0633\u0627\u0645\u062d\u060c \u0648\u064a\u062d\u062b \u0627\u0644\u0625\u0646\u0633\u0627\u0646 \u0639\u0644\u0649 \u0645\u0639\u0627\u0645\u0644\u0629 \u0627\u0644\u0622\u062e\u0631\u064a\u0646 \u0628\u0627\u0644\u0631\u062d\u0645\u0629 \u0648\u0627\u0644\u0639\u062f\u0644.<\/td>\n\u064a\u0648\u0636\u062d \u0627\u0644\u0643\u0627\u062a\u0628 \u0623\u0647\u0645\u064a\u0629 \u0627\u0644\u0625\u062e\u0644\u0627\u0635 \u0641\u064a \u0627\u0644\u0639\u0645\u0644\u060c \u0648\u0623\u0646 \u0627\u0644\u0646\u064a\u0629 \u0627\u0644\u0635\u0627\u062f\u0642\u0629 \u0623\u0633\u0627\u0633 \u0642\u0628\u0648\u0644 \u0627\u0644\u0623\u0639\u0645\u0627\u0644 \u0639\u0646\u062f \u0627\u0644\u0644\u0647.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

2. Motivation and Significance<\/h2>\n

Most contemporary NLP systems rely heavily on large annotated datasets and stable training distributions. In Arabic and other low-resource settings, labelled data are often limited, uneven across dialects and domains, or unavailable altogether. As a result, systems frequently struggle when exposed to noisier real-world text or previously unseen data distributions.<\/p>\n

AraGenre reframes genre classification as a robust generalisation problem rather than a conventional closed-label task. Systems must infer communicative function and stylistic behaviour from limited supervision while generalising to naturally occurring Arabic texts.<\/p>\n

The benchmark is designed to encourage research on hierarchical classification, definition-guided modelling, and robust semantic generalisation under realistic low-resource conditions.<\/p>\n

3. Hierarchical Genre Organisation and Task Design<\/h2>\n

AraGenre is organised as a hierarchical Arabic genre classification task where systems must predict both a broad genre<\/strong> label and a fine-grained specific genre<\/strong> label for each Arabic text segment.<\/p>\n

The benchmark currently includes six broad genres:<\/p>\n