Hugo Gonçalo Oliveira
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Hugo Gonçalo Oliveira has a total of 229 co-authors. Here are the top co-authors by number of shared publications:
- Ana Alves
CISUC/LASI - Centre of Informatics and Systems of the University of Coimbra
14 shared publications Paulo Gomes
Center for Informatics and Systems of the University of Coimbra (CISUC), Portugal
10 shared publications- Catarina Silva
University of Coimbra
9 shared publications - Alexandre Rademaker
IBM Research Brazil
4 shared publications - Bruno Carlos Luís Ferreira
University of Coimbra, DEI, CISUC
4 shared publications - Paulo Gomes
University of Coimbra
4 shared publications
Latest Publications
- 24 June 2026
- Published by Elsevier in Expert Systems with Applications
- Vol. 331
Multi-task specialized expert model for hierarchical aspect-based sentiment analysis in consumer healthcare
- 24 June 2026
- Published by Elsevier in Expert Systems with Applications
- Vol. 331
Multi-task specialized expert model for hierarchical aspect-based sentiment analysis in consumer healthcare
- Open Access
Meta-Judging with Large Language Models: Concepts, Methods, and Challenges
Large language models (LLMs) are evolving fast and are now frequently used as evaluators, in a process typically referred to as LLM-as-a-Judge, which provides quality assessments of model outputs. However, recent research points out significant vulnerabilities in such evaluation, including sensitivity to prompts, systematic biases, verbosity effects, and unreliable or hallucinated rationales. These limitations motivated the development of a more robust paradigm, dubbed LLM-as-a-Meta-Judge. This survey reviews recent advances in meta-judging and organizes the literature, by introducing a framework along six key perspectives: (i) Conceptual Foundations, (ii) Mechanisms of Meta-Judging, (iii) Alignment Training Methods, (iv) Evaluation, (v) Limitations and Failure Modes, and (vi) Future Directions. By analyzing the limitations of LLM-as-a-Judge and summarizing recent advances in meta-judging by LLMs, we argue that LLM-as-a-Meta-Judge offers a promising direction for more stable and trustworthy automated evaluation, while highlighting remaining challenges related to cost, prompt sensitivity, and shared model biases, which must be addressed to advance the next generation of LLM evaluation methodologies.
Open AccessMeta-Judging with Large Language Models: Concepts, Methods, and Challenges
Large language models (LLMs) are evolving fast and are now frequently used as evaluators, in a process typically referred to as LLM-as-a-Judge, which provides quality assessments of model outputs. However, recent research points out significant vulnerabilities in such evaluation, including sensitivity to prompts, systematic biases, verbosity effects, and unreliable or hallucinated rationales. These limitations motivated the development of a more robust paradigm, dubbed LLM-as-a-Meta-Judge. This survey reviews recent advances in meta-judging and organizes the literature, by introducing a framework along six key perspectives: (i) Conceptual Foundations, (ii) Mechanisms of Meta-Judging, (iii) Alignment Training Methods, (iv) Evaluation, (v) Limitations and Failure Modes, and (vi) Future Directions. By analyzing the limitations of LLM-as-a-Judge and summarizing recent advances in meta-judging by LLMs, we argue that LLM-as-a-Meta-Judge offers a promising direction for more stable and trustworthy automated evaluation, while highlighting remaining challenges related to cost, prompt sensitivity, and shared model biases, which must be addressed to advance the next generation of LLM evaluation methodologies.
- 20 November 2025
- Published by Elsevier in Applied Soft Computing
- Vol. 187
Role-aware adapters for dialogue summarization in Seq2Seq models
- 20 November 2025
- Published by Elsevier in Applied Soft Computing
- Vol. 187
Role-aware adapters for dialogue summarization in Seq2Seq models
- Open Access
- 13 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 200001-200014
UnHIDE: A Novel Framework for Unsupervised Human-Interpretable Dialogue Exploration
Dialogue systems are increasingly central to applications in customer service, virtual assistance, and beyond, generating vast amounts of conversational data. While these systems have advanced with the exploitation of large language models (LLMs), they still face key limitations, some, in fact, strengthened by the black-box nature of such models, including the lack of feedback mechanisms and the absence of effective solutions for human-in-the-loop interaction and iterative improvement. As a result, understanding, refining, and debugging dialogue behavior remains a major challenge. To address this, we introduce UnHIDE, a novel, unsupervised framework for Human-Interpretable Dialogue Exploration. UnHIDE is designed to support human understanding of large collections of dialogues by surfacing interpretable structures and trends. It operates in three stages: (1) utterance clustering to group semantically similar dialogue turns, (2) flow discovery to build dialogue trajectories based on these clusters, and (3) the computation of interpretable metrics to analyze flow complexity, sentiment progression, and response times.We evaluate UnHIDE using a newly-created, automatically-generated, task-oriented dialogue dataset, where dialogue length, sentiment dynamics, and timing are systematically varied. Our results show that UnHIDE reliably captures these variations and provides actionable insights into dialogue structure and quality. By enabling transparent, human-interpretable analysis of dialogue without supervision, UnHIDE offers a powerful tool for diagnosing and improving dialogue systems. It not only fills a critical gap in feedback and interpretability, but also lays the groundwork for incorporating human-in-the-loop practices into future conversational Artificial Intelligence (AI) development.
- 13 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 200001-200014
Open AccessUnHIDE: A Novel Framework for Unsupervised Human-Interpretable Dialogue Exploration
Dialogue systems are increasingly central to applications in customer service, virtual assistance, and beyond, generating vast amounts of conversational data. While these systems have advanced with the exploitation of large language models (LLMs), they still face key limitations, some, in fact, strengthened by the black-box nature of such models, including the lack of feedback mechanisms and the absence of effective solutions for human-in-the-loop interaction and iterative improvement. As a result, understanding, refining, and debugging dialogue behavior remains a major challenge. To address this, we introduce UnHIDE, a novel, unsupervised framework for Human-Interpretable Dialogue Exploration. UnHIDE is designed to support human understanding of large collections of dialogues by surfacing interpretable structures and trends. It operates in three stages: (1) utterance clustering to group semantically similar dialogue turns, (2) flow discovery to build dialogue trajectories based on these clusters, and (3) the computation of interpretable metrics to analyze flow complexity, sentiment progression, and response times.We evaluate UnHIDE using a newly-created, automatically-generated, task-oriented dialogue dataset, where dialogue length, sentiment dynamics, and timing are systematically varied. Our results show that UnHIDE reliably captures these variations and provides actionable insights into dialogue structure and quality. By enabling transparent, human-interpretable analysis of dialogue without supervision, UnHIDE offers a powerful tool for diagnosing and improving dialogue systems. It not only fills a critical gap in feedback and interpretability, but also lays the groundwork for incorporating human-in-the-loop practices into future conversational Artificial Intelligence (AI) development.
- Open Access
- 29 September 2025
- Published by Sociedade Brasileira de Computação
- p. 206-219
Improving Pun Detection with an Ensemble of Traditional Machine Learning Methods
Humor is a remarkably complex emotional process, defined as any object or event that causes laughter or amusement or is considered funny. Therefore, recognizing humor is considered one of the most challenging tasks in Natural Language Processing. In this paper, we approached the pun detection task for the Portuguese language. Puns are a form of wordplay that exploits multiple meanings of a term or similar-sounding words to create an intended humorous or rhetorical effect. Our strategy is straightforward: we trained and evaluated an ensemble learning approach of traditional machine learning models on PUNTUGUESE, a recent corpus of Portuguese puns. With this, we outperformed a BERT-based model by 11 p.p. in accuracy and achieved state-of-the-art results. More than that, we performed a detailed error analysis and found that our approach has limitations in identifying puns that contain neologisms.
- 29 September 2025
- Published by Sociedade Brasileira de Computação
- p. 206-219
Open AccessImproving Pun Detection with an Ensemble of Traditional Machine Learning Methods
Humor is a remarkably complex emotional process, defined as any object or event that causes laughter or amusement or is considered funny. Therefore, recognizing humor is considered one of the most challenging tasks in Natural Language Processing. In this paper, we approached the pun detection task for the Portuguese language. Puns are a form of wordplay that exploits multiple meanings of a term or similar-sounding words to create an intended humorous or rhetorical effect. Our strategy is straightforward: we trained and evaluated an ensemble learning approach of traditional machine learning models on PUNTUGUESE, a recent corpus of Portuguese puns. With this, we outperformed a BERT-based model by 11 p.p. in accuracy and achieved state-of-the-art results. More than that, we performed a detailed error analysis and found that our approach has limitations in identifying puns that contain neologisms.
- Open Access
- 25 September 2025
- Published by Wiley in Expert Systems
- Vol. 42 (11)
Twitter and Sentiment Analysis for Wildfire Heat Mapping
Nowadays, automated intelligent systems play an increasingly vital role in aiding decision‐making processes across various fields. Firefighting represents a crucial area where accurate information gathering is paramount for efficient resource allocation. Social media platforms as Twitter (or X) have emerged as valuable sources of real‐time data, often referred to as ‘citizen science’, offering additional insights alongside traditional data sources. In this work, we introduce a novel pipeline that leverages Natural Language Processing (NLP) techniques and Twitter data, utilising transformer models to identify and monitor wildfire incidents. Expanding on this approach, we incorporate sentiment analysis to provide deeper insights into public perceptions and emotions related to fire events. Additionally, we present visual representations of geographic data through heat mapping, potentially aiding firefighters in making informed decisions. By integrating advanced NLP techniques with social media data, our approach presents a promising strategy for enhancing wildfire management efforts.
- 25 September 2025
- Published by Wiley in Expert Systems
- Vol. 42 (11)
Open AccessTwitter and Sentiment Analysis for Wildfire Heat Mapping
Nowadays, automated intelligent systems play an increasingly vital role in aiding decision‐making processes across various fields. Firefighting represents a crucial area where accurate information gathering is paramount for efficient resource allocation. Social media platforms as Twitter (or X) have emerged as valuable sources of real‐time data, often referred to as ‘citizen science’, offering additional insights alongside traditional data sources. In this work, we introduce a novel pipeline that leverages Natural Language Processing (NLP) techniques and Twitter data, utilising transformer models to identify and monitor wildfire incidents. Expanding on this approach, we incorporate sentiment analysis to provide deeper insights into public perceptions and emotions related to fire events. Additionally, we present visual representations of geographic data through heat mapping, potentially aiding firefighters in making informed decisions. By integrating advanced NLP techniques with social media data, our approach presents a promising strategy for enhancing wildfire management efforts.
- 25 September 2025
- Published by Elsevier in Expert Systems with Applications
- Vol. 299
Reasoning or not? A comprehensive evaluation of reasoning LLMs for dialogue summarization
- 25 September 2025
- Published by Elsevier in Expert Systems with Applications
- Vol. 299
Reasoning or not? A comprehensive evaluation of reasoning LLMs for dialogue summarization
- 16 September 2025
- Published by Springer Nature
- p. 435-447
Text Rewriting with Transformers for Humor Generation in Portuguese
- 16 September 2025
- Published by Springer Nature
- p. 435-447
Text Rewriting with Transformers for Humor Generation in Portuguese
- 16 September 2025
- Published by Springer Nature
- p. 249-261
Prompting LLMs for Relation Classification in Portuguese: Is It Worth It?
- 16 September 2025
- Published by Springer Nature
- p. 249-261
Prompting LLMs for Relation Classification in Portuguese: Is It Worth It?