18th-century Fauna and Flora: from Named Entities to the problems of standardization

2026-04-12 · Source: Paper Index on ACL Anthology · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Intermediate, quick

Summary

This article details the challenges and processes involved in preparing 18th-century Portuguese historical sources, specifically those concerning fauna and flora, for computational analysis. The authors initially anticipated low lexical ambiguity due to the specialized domain but encountered significant variation. The work outlines a roadmap for orthographic normalization, describes the creation of an annotated corpus of Named Entities, and discusses the problems arising from lexical variation within these specialized thesauri. The study aims to contribute to the understanding of historical source normalization and emphasize the importance of robust practices in this field.

Key takeaway

For research scientists working with historical linguistic data, you should anticipate significant lexical variation even in specialized domains. Your computational processing pipelines will benefit from a well-defined orthographic normalization process and a carefully constructed annotated corpus of Named Entities to accurately handle historical language nuances.

Key insights

Lexical variation in historical specialized texts presents significant challenges for computational processing.

Principles

• Specialized domains do not guarantee low lexical ambiguity.
• Orthographic normalization is critical for historical texts.

Method

The method involves creating an annotated Named Entity corpus and developing an orthographic normalization roadmap to address lexical variation in historical Portuguese texts.

In practice

• Annotate historical corpora for Named Entities.
• Develop normalization roadmaps for lexical variation.

Topics

• 18th-century Portuguese
• Historical Sources
• Fauna and Flora
• Named Entity Recognition
• Orthographic Normalization

Best for: Research Scientist, AI Scientist, NLP Engineer

Related on AIssential

• Analysing LLMs for spelling normalization of 18th century Portuguese — Specialized LLMs significantly outperform multilingual models for historical language normalization.
• Curation of a Palaeohispanic Dataset for Machine Learning — A new structured dataset enables machine learning analysis of under-resourced Palaeohispanic languages.
• Semantic marking of named entities in The Lusiads — Hybrid NER workflows improve accuracy for historical literary texts, overcoming limitations of models trained on modern corpora.
• Taxonomies, All the Way Down — Taxonomies, a pillar of library science, provide explicit, auditable structure that significantly enhances LLM accuracy and reliability.
• CLEF HIPE-2026: Evaluating Accurate and Efficient Person-Place Relation Extraction from Multilingual Historical Texts — HIPE-2026 evaluates person-place relation extraction in historical texts, emphasizing accuracy, efficiency, and generalization.
• Modelling the Morphology of Verbal Paradigms: A Case Study in the Tokenization of Turkish and Hebrew — Tokenization strategy critically impacts Transformer models' ability to represent complex morphology, especially in non-concatenative languages.

Open in AIssential →

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