A Lexicon-Grammar of Brazilian Portuguese Predicative Adjectives

2026-04-12 · Source: Paper Index on ACL Anthology · Field: Technology & Digital — Artificial Intelligence & Machine Learning · Depth: Expert, short

Summary

Ryan Martinez, Jorge Baptista, and Oto Araújo Vale developed a syntactic lexicon for Brazilian Portuguese predicative adjectives not derived from verbs. This resource, presented at PROPOR 2026, was built by selecting 3,161 lexical items from the 7,000 most frequent adjectives in a large web corpus. Each selected adjective was annotated with 36 syntactic properties, including argument structure, copular verbs, prepositions, transformations like raising and nominalization, and semantic roles. The properties were defined using both introspection and corpus evidence, resulting in a machine-readable lexicon designed to support computational processing of Brazilian Portuguese.

Key takeaway

For research scientists developing natural language processing tools for Brazilian Portuguese, this new lexicon offers a structured, machine-readable resource. You can integrate this detailed syntactic information on predicative adjectives to enhance parsing accuracy, improve semantic role labeling, and refine grammar analysis in your models, particularly for adjectives not directly derived from verbs.

Key insights

A new machine-readable lexicon for Brazilian Portuguese predicative adjectives was created with 36 syntactic properties.

Principles

• Corpus evidence validates linguistic introspection.
• Syntactic properties define adjective behavior.

Method

Selected 3,161 adjectives from 7,000 frequent ones in a web corpus, then annotated them with 36 syntactic properties derived from introspection and corpus evidence.

In practice

• Use for NLP tasks in Brazilian Portuguese.
• Integrate into grammar checkers.

Topics

• Brazilian Portuguese
• Predicative Adjectives
• Syntactic Lexicon
• Lexical Annotation
• Argument Structure

Best for: Research Scientist, AI Scientist, NLP Engineer

Related on AIssential

• Lexicon-Grammar Web — A web-based interface provides interactive access to a large-scale European Portuguese grammatical database.
• Retrato_Cantado: Criação e Análise de um Corpus para Representações de Identidade em Letras de Músicas Brasileiras — Lexical-syntactic patterns effectively identify predicative sentences in song lyrics for identity representation analysis.
• A Scalable Tool for Measuring Manner and Result Verbs in Developmental Language Research — A RoBERTa classifier, trained on LLM-generated annotations, accurately distinguishes manner and result verbs at scale.
• AfriSUD: A Dependency Treebank Collection for Evaluating Models on African Languages — AfriSUD provides the first large-scale, native-speaker verified dependency treebanks for nine diverse African languages, revealing a significant syntax gap in current NLP models.
• Datasets for Verb Alternations across Languages: BLM Templates and Data Augmentation Strategies — This work addresses the underexplored ability of large language models (LLMs) to capture cross-sentence paradigmatic patterns, specifically verb alternations.
• Fully Automated Identification of Lexical Alignment and Preference-Stage Shifts in Large Language Models — Automated metrics can identify and quantify lexical overuse and preference learning shifts in LLMs without manual curation.

Open in AIssential →

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