Evaluating Contextualized Embeddings for Topic Modeling in Public Bidding Domain

Hott, Henrique R.; Silva, Mariana O.; Oliveira, Gabriel P.; Brandão, Michele A.; Lacerda, Anisio; Pappa, Gisele

doi:10.1007/978-3-031-45392-2_27

Cited by 3 publications

(1 citation statement)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For instance, Coelho et al [2022] and address the document classification problem, explicitly focusing on the Brazilian legal and public bidding domains, respectively. With a different task, Hott et al [2023] explore the potential of BERT-based models to generate clusters that effectively encapsulate the underlying topics within bidding data. These efforts represent the growing adoption of advanced computational techniques in government document analysis.…”

Section: Government Documents Processingmentioning

confidence: 99%

Overpricing Analysis in Brazilian Public Bidding Items

Silva,

Costa,

Gomide

et al. 2024

JIS

View full text Add to dashboard Cite

Analyzing overpricing in public bidding items is essential for government agencies to detect signs of fraud in acquiring public goods and services. In this context, this paper presents two main contributions: a methodology for processing and standardizing bid item descriptions and a statistical approach for overpricing detection using the interquartile range. We evaluated a comparative analysis of three distinct grouping strategies, each emphasizing different facets of the item description standardization process. Furthermore, to gauge the efficacy of both proposed approaches, we leveraged a ground-truth dataset for a thorough evaluation containing quantitative and qualitative analyses. Overall, our findings suggest that the evaluated strategies are promising for identifying potential irregularities within public bidding processes.

show abstract

Section: Government Documents Processingmentioning

confidence: 99%

Overpricing Analysis in Brazilian Public Bidding Items

Silva,

Costa,

Gomide

et al. 2024

JIS

View full text Add to dashboard Cite

show abstract

Large Language Models in Biomedical and Health Informatics: A Review with Bibliometric Analysis

Yu,

Fan,

et al. 2024

J Healthc Inform Res

View full text Add to dashboard Cite

Evaluating Domain-adapted Language Models for Governmental Text Classification Tasks in Portuguese

Silva,

Oliveira,

Costa

et al. 2024

Anais Do XXXIX Simpósio Brasileiro De Banco De Dados (SBBD 2024)

View full text Add to dashboard Cite

Domain-adaptive pre-training (DAPT) is a technique in natural language processing (NLP) that tailors pre-trained language models to specific domains, enhancing their performance in real-world applications. In this paper, we evaluate the effectiveness of DAPT in governmental text classification tasks, exploring how different factors, such as target domain dataset, pre-trained model language composition, and dataset size, impact model performance. We systematically vary these factors, creating distinct domain-adapted models derived from BERTimbau and LaBSE. Our experimental results reveal that selecting appropriate target domain datasets and pre-training strategies can notably enhance the performance of language models in governmental tasks.

show abstract

Evaluating Contextualized Embeddings for Topic Modeling in Public Bidding Domain

Cited by 3 publications

References 16 publications

Overpricing Analysis in Brazilian Public Bidding Items

Overpricing Analysis in Brazilian Public Bidding Items

Large Language Models in Biomedical and Health Informatics: A Review with Bibliometric Analysis

Evaluating Domain-adapted Language Models for Governmental Text Classification Tasks in Portuguese

Contact Info

Product

Resources

About