The Quran
and poetry.
Explore the difference.
Compare the Quran with Arabic poetry through word frequencies and linguistic features. See how their samples separate, or compare poetry from different periods using the same features.
Compare Quran & poetry ↗Explore texts across time.
Stylometry Tool is an interactive research website for comparing Arabic poetry from different historical periods with the Quran. Comparisons use the historical folded spelling of the corpus. Feature frequencies, hierarchical clustering (HC), and principal component analysis (PCA) offer complementary views of the selected texts.
The available poetry groups are Pre-Islamic, Transitional poets (Mukhadramun), Umayyad, Abbasid, Andalusian, Fatimid, Ayyubid, Mamluk, Ottoman, and Modern. Choose which eras to include, the number of samples, and the length of each sample. In the experiment lab, Quran, Pre-Islamic poetry, and Modern poetry are selected by default; you can also compare poetry groups with one another.
The experiment lab defaults to 10 samples from each selected group, with 7,000 words per sample, including the Quran. Longer passages and more examples provide a broader basis for estimating feature frequencies and can make recurring patterns clearer. Samples are drawn from non-overlapping passages; groups with less text supply fewer complete samples.
Use the Feature Set dropdown to represent the samples as most frequent words, word bigrams, or character sequences of 1, 2, 3, 5, or 7 characters. Explore their similarities in the Cluster tree and their distribution in the PCA map (principal component analysis). Changing the groups, sampling settings, or feature set automatically updates the results, so you can see how each choice affects the patterns.
The Holy Quran and the challenge to poets — hierarchical clustering
Watch the explainer with English captions, then explore the text comparisons in the experiment lab below. Use the CC menu to turn captions on or off.
Quran source: Tanzil Project — Simple Clean 1.1
Compare texts across groups.
Compare Quran with poetry, or select poetry groups to compare across periods.
Reading the corpus…
Explore your texts with HC and PCA.
Choose a method to explore the same samples from a different perspective.
Quran vs poetry: top 100 most frequent words
Features are listed by total frequency across all selected samples. The Top features setting controls the list size (100 by default); constant features are excluded. Rates are per 10,000 feature events, with standardized differences shown alongside.
| Feature | Quran rate | Poetry rate | Standardized difference | Examples |
|---|
Does the result hold across features?
Reuses your last completed experiment’s texts and settings. Differences reveal sensitivity to feature choice.
How to read these results
Quran & poetry passage clusters
Each point is an equal-length passage. One shared model compares all selected groups; era labels are added after clustering.
Building shared passage embeddings and fitting clusters…
Drag to rotate and scroll to zoom. Color by era compares sources; Color by cluster shows the fitted groups. Select a point to see its nearest passages. PCA, t-SNE and UMAP are available in the left panel.
Use the search box on the right to find passages by their opening excerpt. Enter Arabic text only.
Cluster composition & validation
How to read these clusters
By default, each group contributes 30 separate passages of 2,000 words. All groups use the same number of passages, reduced if a source is shorter.
The tool compares letter patterns or word frequencies and groups similar passages without using group names. Spelling is standardized before comparison.
Clusters show similarities and differences in the selected texts. Try other settings to see how consistent the pattern is.
Corpus & text downloads
Browse the Quran and poetry used in this tool. Download the text files without signing in.
Loading corpus information…
Contact
Dr Ahmed Ibrahim Ahmed Omer
University of York, UK
Email ahmed.omer@york.ac.uk aiomer@hotmail.comWe make this tool available to researchers to analyse texts from these historical periods and enrich the field through further scholarly research. If you use it in a research paper, please cite it in the text and include it in your reference list using the format below.
Reference template (APA 7)
Omer, A. I. A. (n.d.). Stylometry Tool: Quran & poetry [Computer software]. [Public URL of the tool]
In-text citation: (Omer, n.d.)
Replace the URL placeholder with the public link to the tool. Use the release year and version number when available; n.d. means no date.