Weslley Emmanuel Martins Lima

Prof. Weslley Emmanuel Martins Lima


Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
9 December 2024 in Data in Brief
Verified
Journal Article
Open access
WLWeslley Emmanuel Martins Lima
Weslley Emmanuel Martins Lima
VSVictor Silva
Victor Silva
JSJasson Silva
Jasson Silva
RLRicardo Lira
Ricardo Lira
APAnselmo Paiva
Anselmo Paiva
Digital transformation has significantly impacted public procurement, improving operational efficiency, transparency, and competition. This transformation has allowed the automation of data analysis and oversight in public administration. Public procurement involves various stages and generates a multitude of documents. However, experts manually analyze these unstructured textual documents, which are time-consuming and inefficient. To address this issue, we introduce BidCorpus, a novel and comprehensive dataset consisting of thousands of documents related to public procurement, specifically bidding notices from Brazilian public websites. The dataset was labeled using weak supervision techniques, manual labeling, and BERT-based language models. Models trained with these annotated data showed promising results, with metrics greater than 80 % in various experiments. The models could also tolerate intentional changes made to bidding notices to evade fraud detection. All the resources from this work are publicly available, including the documents, pre-processing scripts, and training and evaluation of the models. We expect the dataset and its labels to be of great value to researchers working on public procurement problems.

22 March 2025 in Electronics
Verified
Journal Article
Open access
PDF file
HSHélcio de Abreu Soares
Hélcio de Abreu Soares
RMRaimundo Moura
Raimundo Moura
VMVinícius Ponte Machado
Vinícius Ponte Machado
Anselmo C PaivaAnselmo C Paiva
Anselmo C Paiva
WLWeslley Emmanuel Martins Lima
Weslley Emmanuel Martins Lima
RVRodrigo Veras
Rodrigo Veras
Artificial Intelligence (AI) models, including deep learning and rule-based approaches, often function as black boxes, limiting transparency and increasing uncertainty in decisions. This study addresses spurious correlations, defined as associations between patterns and classes that do not reflect causal relationships, affecting AI models’ reliability and applicability. In Natural Language Processing (NLP), these correlations lead to inaccurate predictions, biases, and challenges in model generalization. We propose a method that employs Explainable Artificial Intelligence (XAI) techniques to detect spurious patterns in textual datasets for binary classification tasks. The method applies the K-means algorithm to cluster patterns and interprets them based on their distance from centroids. It hypothesizes that patterns farther from the centroids are more likely to be spurious than those closer to them. We apply the method to public procurement datasets from the Court of Auditors of Piauí (TCE-PI) using models based on Support Vector Machine (SVM) and Logistic Regression with text representations from TFIDF and Word Embeddings, as well as a BERT model. The analysis is extended to the IMDB movie review dataset to evaluate generalizability. The results support the hypothesis that patterns farther from centroids exhibit higher spuriousness potential and demonstrate the clustering’s consistency across models and datasets. The method operates independently of the techniques used in its stages, enabling the automatic detection and quantification of spurious patterns without prior human intervention.

20 January 2026 in Journal of the Brazilian Computer Society
Verified
Journal Article
Open access
PDF file
WLWeslley Emmanuel Martins Lima
Weslley Emmanuel Martins Lima
VSVictor Ribeiro Da Silva
Victor Ribeiro Da Silva
JSJasson Carvalho Da Silva
Jasson Carvalho Da Silva
Ricardo RabeloRicardo Rabelo
Ricardo Rabelo
Anselmo C PaivaAnselmo C Paiva
Anselmo C Paiva
Deep learning models excel in various tasks but require extensive annotated data for supervised learning. In NLP, limited annotated data hinders deep learning. Self-supervised pretraining addresses this by training models on unlabeled text to learn useful representations. Domain-specific pretraining is crucial for good performance in downstream tasks. Although pretrained BERT models exist for legal documents in some languages, none target public procurement documents in Portuguese. Public procurement documents have terminology that is not found in existing models. In this paper, we propose HelBERT, a BERT-based model pretrained on a large corpus of public procurement documents in the Brazilian Portuguese language, including laws, tender notices, and contracts. The experimental results demonstrate that HelBERT outperforms other models in all analyses. HelBERT surpasses models such as BERTimbau and JurisBERT in classification tasks by achieving improvements of 5% and 4% in the F1 Score, respectively. Furthermore, the model achieves gains that exceed 3% in semantic similarity tasks compared to the baseline models. Moreover, despite using a GPU with reduced memory and processing resources, the proposed approach achieves superior results with fewer and more efficient training epochs than the baseline models. These findings underscore the effectiveness of the proposed model in addressing NLP tasks within the public procurement domain.

11 September 2024
Verified
Conference Paper
Open access
PDF file
JSJasson Carvalho da Silva
Jasson Carvalho da Silva
Ricardo RabeloRicardo Rabelo
Ricardo Rabelo
WLWeslley Emmanuel Martins Lima
Weslley Emmanuel Martins Lima
VAVitor Augusto Correa Cortez Almeida
Vitor Augusto Correa Cortez Almeida
O estudo visa avaliar e implementar técnicas de agregação de características em modelos hierárquicos para lidar com textos longos no domínio jurídico. Modelos baseados em transformers, embora eficazes em vários problemas de Processamento de Linguagem Natural, enfrentam dificuldades devido à sua complexidade quadrática, especialmente com textos que excedem 512 tokens. Para superar esses desafios, o estudo propõe investigar métodos como modelos hierárquicos, estratégias de divisão de texto, atenção local e esparsa, com o objetivo de aprimorar o processamento e a análise desses textos, fornecendo insights valiosos para a compreensão e uso eficaz em contextos jurídicos.