Gendered Stylistic Variation in Brazilian Portuguese Google Play Reviews: A Large-Scale Study

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

Summary

A large-scale study analyzed gender-associated stylistic variations in 76.7 million Brazilian Portuguese Google Play reviews from 96 apps between 2011 and 2025. Researchers inferred binary gender from first names using IBGE name frequencies, yielding 22.25 million high-confidence labels. The findings indicate that reviews associated with women exhibit approximately 60% higher paralinguistic expressivity, characterized by increased emoji density, lengthening, and punctuation. Conversely, lexical diversity, measured by MTLD, remained nearly identical across gender groups. While overall ratings were predominantly positive, men contributed a relatively higher proportion of 1-star reviews, whereas women submitted more 5-star reviews. This research offers insights into digital sociolinguistic behavior within the Brazilian context.

Key takeaway

For NLP engineers developing sentiment analysis or user profiling models for Brazilian Portuguese, you should account for gender-associated stylistic variations. Specifically, integrate paralinguistic features like emoji density and punctuation into your models, as these show significant differences and could improve the accuracy of sentiment or demographic predictions, especially when analyzing user reviews.

Key insights

Gender-associated stylistic differences exist in Brazilian Portuguese Google Play reviews, particularly in paralinguistic expressivity.

Principles

• Name-based inference can label gender in large datasets.
• Paralinguistic features vary more by gender than lexical diversity.

Method

Binary gender was inferred from first names in 76.7M Google Play reviews using IBGE name frequencies to identify stylistic variations.

In practice

• Analyze paralinguistic features for demographic insights.
• Consider name-based gender inference for large text corpora.

Topics

• Gendered Stylistic Variation
• Brazilian Portuguese
• Google Play Reviews
• Sociolinguistics
• Paralinguistic Expressivity

Best for: NLP Engineer, AI Scientist, Research Scientist

Related on AIssential

• Gender Identification in Brazilian Portuguese Product Reviews: A Comparative Study of Classical Models, BERT, and LLMs — Gender identification in Brazilian Portuguese Amazon reviews shows BERT outperforming LLMs, with performance varying by author gender and product domain.
• Compression-based Language Complexity under Register Variation in Portuguese — Compression-based metrics effectively measure linguistic complexity and register variation across languages.
• Machine learning and digital pragmatics: Which word category influences emoji use most? — MARBERT model shows promise in predicting Arabic emoji use, but needs further refinement for dialectal nuances.
• As Easy as Rocket Science: Assessing the Ability of Large Language Models to Interpret Negation in Figurative Language — LLMs struggle with combined negation and figurative language, with prompt style significantly impacting performance.
• Boss vs. Bloom: Uncovering Gender Bias in Fragrance Descriptions with Python — Fragrance marketing language and pricing systematically reinforce traditional gender stereotypes.
• MORPHOGEN: A Multilingual Benchmark for Evaluating Gender-Aware Morphological Generation — MORPHOGEN evaluates LLM gender-aware morphological generation across French, Arabic, and Hindi, revealing performance gaps and biases.

Open in AIssential →

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