Confidence‑weighted aggregation into a Poet × Concept matrix interpreted as probability distributions
Use of Jensen–Shannon and Kullback–Leibler divergences to quantify individual poetic distinctiveness
Construction of a confidence‑weighted concept co‑occurrence graph
Eigenmood embedding via Laplacian spectral decomposition of the graph
Corpus of 61,573 verses across ten poets, with 22.2 % abstentions highlighting uncertainty
Sensitivity analysis under confidence thresholding and selection‑bias diagnostics treating abstention as a category
Distant‑to‑close workflow that retrieves verse‑level exemplars along Eigenmood axes
Goal of providing a reproducible, auditable digital‑humanities pipeline that retains interpretive rigor
📖 Full Retelling
The researchers Kourosh Shahnazari, Seyed Moein Ayyoubzadeh, and Mohammadali Keshtparvar published a paper titled *Eigenmood Space: Uncertainty‑Aware Spectral Graph Analysis of Psychological Patterns in Classical Persian Poetry* on the preprint server arXiv on 18 February 2026. The paper introduces an uncertainty‑aware computational framework that blends large‑scale automatic multi‑label annotation with probability‑weighted poet–concept matrices, divergence‑based measures of poetic individuality, and a Laplacian spectral embedding called Eigenmood. By propagating per‑verse confidence scores and abstention flags, the authors aim to offer a scalable, reproducible, and auditable digital‑humanities method that preserves interpretive caution while uncovering relationships among psychological concepts in a corpus of 61,573 verses.
🏷️ Themes
Computational linguistics, Digital humanities, Uncertainty quantification, Spectral graph theory, Literary analysis of classical poetry
Entity Intersection Graph
No entity connections available yet for this article.
Deep Analysis
Why It Matters
This study introduces a computational framework that quantifies psychological patterns in classical Persian poetry while explicitly accounting for uncertainty, enabling large‑scale, auditable analysis that preserves interpretive caution. By combining multi‑label annotation, confidence weighting, and spectral graph embeddings, it bridges the gap between close literary reading and scalable digital humanities research.
Context & Background
Classical Persian poetry conveys affective life through metaphor and intertextual convention.
Traditional close reading limits reproducible comparison at scale.
The new framework aggregates confidence‑weighted evidence into a poet‑by‑concept matrix and uses spectral graph analysis to capture relational structure.
What Happens Next
Future work will extend the framework to larger corpora and other literary traditions, integrate it with existing digital humanities platforms, and refine uncertainty handling for broader applications.
Frequently Asked Questions
What is Eigenmood embedding?
It is a vector representation derived from Laplacian spectral decomposition of a confidence‑weighted co‑occurrence graph of psychological concepts.
How is uncertainty handled in the analysis?
Each verse receives a confidence score per label and an abstention flag; these propagate through the pipeline to poet‑level inference.
Can this method be applied to other languages or literary traditions?
Yes, the framework is language‑agnostic and can be adapted to other corpora with similar psychological annotation schemes.
}
Original Source
--> Computer Science > Computation and Language arXiv:2602.16959 [Submitted on 18 Feb 2026] Title: Eigenmood Space: Uncertainty-Aware Spectral Graph Analysis of Psychological Patterns in Classical Persian Poetry Authors: Kourosh Shahnazari , Seyed Moein Ayyoubzadeh , Mohammadali Keshtparvar View a PDF of the paper titled Eigenmood Space: Uncertainty-Aware Spectral Graph Analysis of Psychological Patterns in Classical Persian Poetry, by Kourosh Shahnazari and 2 other authors View PDF HTML Abstract: Classical Persian poetry is a historically sustained archive in which affective life is expressed through metaphor, intertextual convention, and rhetorical indirection. These properties make close reading indispensable while limiting reproducible comparison at scale. We present an uncertainty-aware computational framework for poet-level psychological analysis based on large-scale automatic multi-label annotation. Each verse is associated with a set of psychological concepts, per-label confidence scores, and an abstention flag that signals insufficient evidence. We aggregate confidence-weighted evidence into a Poet $\times$ Concept matrix, interpret each poet as a probability distribution over concepts, and quantify poetic individuality as divergence from a corpus baseline using Jensen--Shannon divergence and Kullback--Leibler divergence. To capture relational structure beyond marginals, we build a confidence-weighted co-occurrence graph over concepts and define an Eigenmood embedding through Laplacian spectral decomposition. On a corpus of 61{,}573 verses across 10 poets, 22.2\% of verses are abstained, underscoring the analytical importance of uncertainty. We further report sensitivity analysis under confidence thresholding, selection-bias diagnostics that treat abstention as a category, and a distant-to-close workflow that retrieves verse-level exemplars along Eigenmood axes. The resulting framework supports scalable, auditable digital-humanities analysis while preservin...
