In Search of Lost Adventure Novels: Supervised Genre Retrieval and Corpus Refinement in Gallica

2026-06-25 · Source: Paper Index on ACL Anthology · Field: Science & Research — Artificial Intelligence & Machine Learning, Data Science & Analytics, Social Sciences & Behavioral Studies · Depth: Intermediate, quick

Summary

This paper addresses the challenge of retrieving adventure novels from Gallica, a large digitized collection of French fiction with unreliable genre metadata. Researchers developed a supervised genre model using a seed list of 101 adventure novels and validated it against 364 independently labeled novels from the Chapitres corpus. The HistGradientBoosting model, utilizing mean paragraph embeddings, achieved 81% external recall despite its small training set. This model was then applied to the 12,176-novel Fictions littde Gallica archive. A subsequent graph-based post-processing step, using a k-nearest-neighbor similarity graph, refined the candidate corpus, producing a more cohesive and homogeneous collection on Gallica, identifying both missed canonical and borderline texts.

Key takeaway

For computational literary historians or digital humanities researchers working with large, uncurated text archives, you should consider a two-stage approach for genre retrieval. First, train a robust supervised classifier like HistGradientBoosting on a small, high-quality seed list. Then, apply graph-based post-processing to refine the initial results, enhancing corpus cohesion and surfacing overlooked texts, rather than solely relying on classification for final curation.

Key insights

Supervised genre modeling combined with graph-based refinement effectively retrieves and curates specific literary genres from large, noisy digital archives.

Principles

• Small training sets can yield strong recall.
• Graph-based correction refines noisy corpora.
• Heuristics complement supervised classification.

Method

A supervised genre classifier (HistGradientBoosting on mean paragraph embeddings) is applied to a large corpus, followed by graph-based post-processing over a k-nearest-neighbor similarity graph for refinement.

In practice

• Use HistGradientBoosting for text classification.
• Apply k-NN graphs for corpus refinement.
• Identify missed canonical texts.

Topics

• Genre Retrieval
• Computational Literary History
• HistGradientBoosting
• Gallica Archive
• Graph-based Post-processing
• k-Nearest-Neighbor

Best for: NLP Engineer, Research Scientist, AI Scientist, Machine Learning Engineer

Related on AIssential

• Rethinking Negative Sampling for Knowledge Distillation in Retrieval, A Distillation Recipe for Small Language Model Search Agents, and More! — Effective information retrieval advancements often stem from rethinking fundamental assumptions and optimizing data representation or model behavior.
• Recognising non-named spatial entities in literary texts: a novel spatial entities classifier — A novel deep learning classifier identifies non-named spatial entities in Swiss-German novels using BERT and Prodigy.
• HistoriQA-ThirdRepublic: Multi-Hop Question Answering Corpus for Historical Research, Parliamentary Debates from the French Third Republic (1870-1940) — HistoriQA-ThirdRepublic bridges NLP benchmarks and historical scholarship through complex multi-hop QA.
• GraphLit: Learning Text-Enriched Dynamic Character Network Representations for Literary Study — GraphLit integrates dynamic character networks with textual context via self-supervised learning to enhance literary analysis.
• ScholarSum: Student-Teacher Abstractive Summarization via Knowledge Graph Reasoning and Reflective Refinement — ScholarSum integrates hierarchical knowledge graphs and iterative student-teacher refinement for fluent, factually faithful scientific summarization.
• DocRetriever: A Plug-and-Play Framework for Multimodal Document Retrieval with Comprehensive Benchmark — DocRetriever improves multimodal document retrieval via layout-aware sparse embeddings and a generalizable, reasoning-augmented reranker.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Paper Index on ACL Anthology.