Slide
Centre Interdisciplinaire
de Recherche et d’Innovation
en Cybersécurité et Société
de Recherche et d’Innovation
en Cybersécurité et Société
1.
Gapski, M. C. B.; Kawai, V. A. S.; Leticio, G. R.; Valem, L. P.; Pedronette, D. C. G.; Allili, M. S.
Graph Neural Networks for Semi-Supervised Image Classification with Multi-Feature Aggregation Journal Article
In: Journal of the Brazilian Computer Society, vol. 32, no. 1, pp. 1731–1754, 2026, ISSN: 01046500 (ISSN).
Abstract | Links | BibTeX | Tags: Classification accuracy, Convolution, Convolutional neural networks, extraction, Feature extraction, Feature extractor, Feature Fusion, Feature representation, Features fusions, Graph Neural Networks, Graph representation, Graph structures, Graph theory, Graphic methods, Image classification, Labeled data, Rank Aggregation, Semi-supervised, Semi-supervised image classification, Supervised image classifications, Textures
@article{gapskiGraphNeuralNetworks2026,
title = {Graph Neural Networks for Semi-Supervised Image Classification with Multi-Feature Aggregation},
author = {M. C. B. Gapski and V. A. S. Kawai and G. R. Leticio and L. P. Valem and D. C. G. Pedronette and M. S. Allili},
url = {https://www.scopus.com/pages/publications/105045179876?origin=resultslist},
doi = {10.5753/jbcs.2026.5880},
issn = {01046500 (ISSN)},
year = {2026},
date = {2026-01-01},
journal = {Journal of the Brazilian Computer Society},
volume = {32},
number = {1},
pages = {1731–1754},
publisher = {Brazilian Computing Society},
abstract = {Feature extraction involves the identification and extraction of salient characteristics or patterns, including edges, textures, shapes, and color attributes. Contemporary feature extractors predominantly leverage deep learning architectures, such as Convolutional Neural Networks (CNNs) and Vision Transformers (VITs). The availability of diverse feature extractors in the literature provides a wide range of feature representations. Features extracted from an image depend on the specific application, the chosen extractor, and its configuration. Therefore, integrating complementary information by combining distinct extractors offers a promising way to enhance performance. Graph Neural Networks (GNNs), particularly Graph Convolutional Networks (GCNs), have emerged as powerful and widely adopted approaches for semi-supervised image classification, as they effectively leverage both labeled and unlabeled data while exploiting the underlying graph structures that capture relationships among samples. This study proposes a novel approach for GNNs in scenarios where labeled data is scarce, by integrating diverse sets of feature and graph representations derived from various extractors in classification scenarios. Experimental investigations were conducted, encompassing combinations of distinct feature and graph extractors, as well as rank aggregation strategies. The primary contributions of this work are underscored by the experimental findings, which demonstrate that the strategic combination of feature and graph representations, coupled with the application of manifold learning for graph processing, leads to significant improvements in classification accuracy across the majority of experimental conditions. Furthermore, the utilization of rank aggregation techniques to integrate features from different extractors was shown to enhance classification accuracy. © 2026, Brazilian Computing Society. All rights reserved.},
keywords = {Classification accuracy, Convolution, Convolutional neural networks, extraction, Feature extraction, Feature extractor, Feature Fusion, Feature representation, Features fusions, Graph Neural Networks, Graph representation, Graph structures, Graph theory, Graphic methods, Image classification, Labeled data, Rank Aggregation, Semi-supervised, Semi-supervised image classification, Supervised image classifications, Textures},
pubstate = {published},
tppubtype = {article}
}
Feature extraction involves the identification and extraction of salient characteristics or patterns, including edges, textures, shapes, and color attributes. Contemporary feature extractors predominantly leverage deep learning architectures, such as Convolutional Neural Networks (CNNs) and Vision Transformers (VITs). The availability of diverse feature extractors in the literature provides a wide range of feature representations. Features extracted from an image depend on the specific application, the chosen extractor, and its configuration. Therefore, integrating complementary information by combining distinct extractors offers a promising way to enhance performance. Graph Neural Networks (GNNs), particularly Graph Convolutional Networks (GCNs), have emerged as powerful and widely adopted approaches for semi-supervised image classification, as they effectively leverage both labeled and unlabeled data while exploiting the underlying graph structures that capture relationships among samples. This study proposes a novel approach for GNNs in scenarios where labeled data is scarce, by integrating diverse sets of feature and graph representations derived from various extractors in classification scenarios. Experimental investigations were conducted, encompassing combinations of distinct feature and graph extractors, as well as rank aggregation strategies. The primary contributions of this work are underscored by the experimental findings, which demonstrate that the strategic combination of feature and graph representations, coupled with the application of manifold learning for graph processing, leads to significant improvements in classification accuracy across the majority of experimental conditions. Furthermore, the utilization of rank aggregation techniques to integrate features from different extractors was shown to enhance classification accuracy. © 2026, Brazilian Computing Society. All rights reserved.
2.
Souza, J. V.; Amamou, H.; Chen, R.; Salari, E.; Gubelmann, R.; Niklaus, C.; Serpa, T.; Lima, M. M. F.; Pinto, P. T.; Kshirsagar, S.; Davoust, A.; Handschuh, S.; Avila, A. R.
Cross-Lingual Keyword Extraction for Pesticide Terminology in Brazilian Portuguese and English Journal Article
In: Journal of the Brazilian Computer Society, vol. 31, no. 1, pp. 973–990, 2025, ISSN: 01046500 (ISSN).
Abstract | Links | BibTeX | Tags: Agriculture, BERT embedding, BERT embeddings, Cross-lingual, Embeddings, extraction, Food consumption, Keywords extraction, Labelings, Low resource languages, Multilingual extraction, Pesticides, Technical terms, Terminology, Word alignment
@article{de_souza_cross-lingual_2025,
title = {Cross-Lingual Keyword Extraction for Pesticide Terminology in Brazilian Portuguese and English},
author = {J. V. Souza and H. Amamou and R. Chen and E. Salari and R. Gubelmann and C. Niklaus and T. Serpa and M. M. F. Lima and P. T. Pinto and S. Kshirsagar and A. Davoust and S. Handschuh and A. R. Avila},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-105019700300&doi=10.5753%2Fjbcs.2025.5815&partnerID=40&md5=85ee75baf4550666a307310cd04d1c83},
doi = {10.5753/jbcs.2025.5815},
issn = {01046500 (ISSN)},
year = {2025},
date = {2025-01-01},
journal = {Journal of the Brazilian Computer Society},
volume = {31},
number = {1},
pages = {973–990},
abstract = {Agriculture plays a crucial role in Brazil’s economy. As the country intensifies its activities in the sector, the use of pesticides also increases. Hence, the risks associated with pesticide-laden food consumption have become a concern for chemistry researchers. An issue affecting regulatory standardization of pesticides in Brazil is the difficulty in translating pesticide names, particularly from English. For example, the word malathion can be translated from English to Portuguese as malatiom or malatião, resulting in inconsistent labeling. This issue extends to the broader problem of translating highly technical terms between languages, in particular for low-resource languages. In this work, we investigate terminological variation in the chemistry of organophosphorus pesticides. Our goal is to study strategies for domain-specific multilingual keyword extraction. To that end, two corpora were built based on pesticide-related scientific documents in Brazilian Portuguese and English, which led to a total of 84 and 210 texts, respectively, representing the low-and high-resource languages in this study. We then assessed 6 methods for keyword extraction: Simple Maths, TF-IDF, YAKE, TextRank, MultipartiteRank, and KeyBERT. We relied on a multilingual contextual BERT embedding to retrieve corresponding pesticide names in the target language. Finetuning was also explored to improve the multilingual representation further. Moreover, we evaluated the use of large language models (LLMs) combined with the recent retrieval-augmented generation (RAG) framework. As a result, we found that the contextual approach, combined with fine-tuning, provided the best results, contributing to enhancing Pesticide Terminology Extraction in a multilingual scenario. © 2025, Brazilian Computing Society. All rights reserved.},
keywords = {Agriculture, BERT embedding, BERT embeddings, Cross-lingual, Embeddings, extraction, Food consumption, Keywords extraction, Labelings, Low resource languages, Multilingual extraction, Pesticides, Technical terms, Terminology, Word alignment},
pubstate = {published},
tppubtype = {article}
}
Agriculture plays a crucial role in Brazil’s economy. As the country intensifies its activities in the sector, the use of pesticides also increases. Hence, the risks associated with pesticide-laden food consumption have become a concern for chemistry researchers. An issue affecting regulatory standardization of pesticides in Brazil is the difficulty in translating pesticide names, particularly from English. For example, the word malathion can be translated from English to Portuguese as malatiom or malatião, resulting in inconsistent labeling. This issue extends to the broader problem of translating highly technical terms between languages, in particular for low-resource languages. In this work, we investigate terminological variation in the chemistry of organophosphorus pesticides. Our goal is to study strategies for domain-specific multilingual keyword extraction. To that end, two corpora were built based on pesticide-related scientific documents in Brazilian Portuguese and English, which led to a total of 84 and 210 texts, respectively, representing the low-and high-resource languages in this study. We then assessed 6 methods for keyword extraction: Simple Maths, TF-IDF, YAKE, TextRank, MultipartiteRank, and KeyBERT. We relied on a multilingual contextual BERT embedding to retrieve corresponding pesticide names in the target language. Finetuning was also explored to improve the multilingual representation further. Moreover, we evaluated the use of large language models (LLMs) combined with the recent retrieval-augmented generation (RAG) framework. As a result, we found that the contextual approach, combined with fine-tuning, provided the best results, contributing to enhancing Pesticide Terminology Extraction in a multilingual scenario. © 2025, Brazilian Computing Society. All rights reserved.
3.
Boudra, N.; Elhajjout, A.; Moudoud, H.; Oujaoura, M.; Jarir, Z.; Houda, Z. A. El
Toward Lightweight IoC Extraction in IoT: The Role of Small Language Models Proceedings Article
In: Int. Congr. Smart Agric. Sustain. Syst., SmartAgri SuSY, Institute of Electrical and Electronics Engineers Inc., 2025, ISBN: 979-833157801-5 (ISBN), (Journal Abbreviation: Int. Congr. Smart Agric. Sustain. Syst., SmartAgri SuSY).
Abstract | Links | BibTeX | Tags: Artificial intelligence, Cyber threats, Data-source, Edge Computing, extraction, Indicator of compromize, Indicators of Compromise, Information retrieval, Information Security, IoT Security, Language model, Malware, Natural languages, Network security, Program processors, Resource Constraint, Resource Constraints, Semantics, Small language model, Small Language Models, Threat Intelligence
@inproceedings{boudraLightweightIoCExtraction2025,
title = {Toward Lightweight IoC Extraction in IoT: The Role of Small Language Models},
author = {N. Boudra and A. Elhajjout and H. Moudoud and M. Oujaoura and Z. Jarir and Z. A. El Houda},
url = {https://www.scopus.com/pages/publications/105037622873?origin=resultslist},
doi = {10.1109/SmartAgriSuSY68475.2025.11466843},
isbn = {979-833157801-5 (ISBN)},
year = {2025},
date = {2025-01-01},
booktitle = {Int. Congr. Smart Agric. Sustain. Syst., SmartAgri SuSY},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {The proliferation of cyber threats necessitates rapid and accurate extraction of Indicators of Compromise (IoCs) from diverse security data sources. While Large Language Models (LLMs) have demonstrated exceptional capabilities in natural language processing and information extraction tasks, their deployment on resource-constrained IoT devices remains challenging due to computational and memory requirements. This paper investigates whether Small Language Models (SLMs) can serve as effective alternatives to LLMs for IoC extraction in IoT environments with limited resources. We present a novel direct LLM-based IoC extraction system leveraging context memory mechanisms for document-wide semantic understanding, specifically designed for deployment on edge computing infrastructure with consumer-grade hardware. Our experimental setup utilizes an RTX 4080 GPU and Ryzen 7 7700X processor running the Ollama framework with GPT-OSS:20B model to evaluate the feasibility of using smaller models instead of resource-intensive LLMs for security tasks. Evaluated on 9 diverse threat intelligence reports spanning different malware families and attack campaigns, the system achieved an average F1 score of 0.62, precision of 0.54, recall of 0.79, and accuracy of 0.85, demonstrating that SLMs can achieve acceptable accuracy levels for IoC extraction while operating within the computational constraints typical of IoT edge deployments. The results suggest that smaller models may provide viable alternatives to large language models for distributed threat intelligence processing in resource-limited environments. © 2025 IEEE.},
note = {Journal Abbreviation: Int. Congr. Smart Agric. Sustain. Syst., SmartAgri SuSY},
keywords = {Artificial intelligence, Cyber threats, Data-source, Edge Computing, extraction, Indicator of compromize, Indicators of Compromise, Information retrieval, Information Security, IoT Security, Language model, Malware, Natural languages, Network security, Program processors, Resource Constraint, Resource Constraints, Semantics, Small language model, Small Language Models, Threat Intelligence},
pubstate = {published},
tppubtype = {inproceedings}
}
The proliferation of cyber threats necessitates rapid and accurate extraction of Indicators of Compromise (IoCs) from diverse security data sources. While Large Language Models (LLMs) have demonstrated exceptional capabilities in natural language processing and information extraction tasks, their deployment on resource-constrained IoT devices remains challenging due to computational and memory requirements. This paper investigates whether Small Language Models (SLMs) can serve as effective alternatives to LLMs for IoC extraction in IoT environments with limited resources. We present a novel direct LLM-based IoC extraction system leveraging context memory mechanisms for document-wide semantic understanding, specifically designed for deployment on edge computing infrastructure with consumer-grade hardware. Our experimental setup utilizes an RTX 4080 GPU and Ryzen 7 7700X processor running the Ollama framework with GPT-OSS:20B model to evaluate the feasibility of using smaller models instead of resource-intensive LLMs for security tasks. Evaluated on 9 diverse threat intelligence reports spanning different malware families and attack campaigns, the system achieved an average F1 score of 0.62, precision of 0.54, recall of 0.79, and accuracy of 0.85, demonstrating that SLMs can achieve acceptable accuracy levels for IoC extraction while operating within the computational constraints typical of IoT edge deployments. The results suggest that smaller models may provide viable alternatives to large language models for distributed threat intelligence processing in resource-limited environments. © 2025 IEEE.
4.
Souza, J. V.; Amamou, H.; Chen, R.; Salari, E.; Gubelmann, R.; Niklaus, C.; Serpa, T.; Lima, M. M. F.; Pinto, P. T.; Kshirsagar, S.; Davoust, A.; Handschuh, S.; Avila, A. R.
Cross-Lingual Keyword Extraction for Pesticide Terminology in Brazilian Portuguese and English Journal Article
In: Journal of the Brazilian Computer Society, vol. 31, no. 1, pp. 973–990, 2025, ISSN: 01046500 (ISSN).
Abstract | Links | BibTeX | Tags: Agriculture, BERT embedding, BERT embeddings, Cross-lingual, Embeddings, extraction, Food consumption, Keywords extraction, Labelings, Low resource languages, Multilingual extraction, Pesticides, Technical terms, Terminology, Word alignment
@article{desouzaCrossLingualKeywordExtraction2025,
title = {Cross-Lingual Keyword Extraction for Pesticide Terminology in Brazilian Portuguese and English},
author = {J. V. Souza and H. Amamou and R. Chen and E. Salari and R. Gubelmann and C. Niklaus and T. Serpa and M. M. F. Lima and P. T. Pinto and S. Kshirsagar and A. Davoust and S. Handschuh and A. R. Avila},
url = {https://www.scopus.com/pages/publications/105019700300?origin=resultslist},
doi = {10.5753/jbcs.2025.5815},
issn = {01046500 (ISSN)},
year = {2025},
date = {2025-01-01},
journal = {Journal of the Brazilian Computer Society},
volume = {31},
number = {1},
pages = {973–990},
publisher = {Brazilian Computing Society},
abstract = {Agriculture plays a crucial role in Brazil’s economy. As the country intensifies its activities in the sector, the use of pesticides also increases. Hence, the risks associated with pesticide-laden food consumption have become a concern for chemistry researchers. An issue affecting regulatory standardization of pesticides in Brazil is the difficulty in translating pesticide names, particularly from English. For example, the word malathion can be translated from English to Portuguese as malatiom or malatião, resulting in inconsistent labeling. This issue extends to the broader problem of translating highly technical terms between languages, in particular for low-resource languages. In this work, we investigate terminological variation in the chemistry of organophosphorus pesticides. Our goal is to study strategies for domain-specific multilingual keyword extraction. To that end, two corpora were built based on pesticide-related scientific documents in Brazilian Portuguese and English, which led to a total of 84 and 210 texts, respectively, representing the low-and high-resource languages in this study. We then assessed 6 methods for keyword extraction: Simple Maths, TF-IDF, YAKE, TextRank, MultipartiteRank, and KeyBERT. We relied on a multilingual contextual BERT embedding to retrieve corresponding pesticide names in the target language. Finetuning was also explored to improve the multilingual representation further. Moreover, we evaluated the use of large language models (LLMs) combined with the recent retrieval-augmented generation (RAG) framework. As a result, we found that the contextual approach, combined with fine-tuning, provided the best results, contributing to enhancing Pesticide Terminology Extraction in a multilingual scenario. © 2025, Brazilian Computing Society. All rights reserved.},
keywords = {Agriculture, BERT embedding, BERT embeddings, Cross-lingual, Embeddings, extraction, Food consumption, Keywords extraction, Labelings, Low resource languages, Multilingual extraction, Pesticides, Technical terms, Terminology, Word alignment},
pubstate = {published},
tppubtype = {article}
}
Agriculture plays a crucial role in Brazil’s economy. As the country intensifies its activities in the sector, the use of pesticides also increases. Hence, the risks associated with pesticide-laden food consumption have become a concern for chemistry researchers. An issue affecting regulatory standardization of pesticides in Brazil is the difficulty in translating pesticide names, particularly from English. For example, the word malathion can be translated from English to Portuguese as malatiom or malatião, resulting in inconsistent labeling. This issue extends to the broader problem of translating highly technical terms between languages, in particular for low-resource languages. In this work, we investigate terminological variation in the chemistry of organophosphorus pesticides. Our goal is to study strategies for domain-specific multilingual keyword extraction. To that end, two corpora were built based on pesticide-related scientific documents in Brazilian Portuguese and English, which led to a total of 84 and 210 texts, respectively, representing the low-and high-resource languages in this study. We then assessed 6 methods for keyword extraction: Simple Maths, TF-IDF, YAKE, TextRank, MultipartiteRank, and KeyBERT. We relied on a multilingual contextual BERT embedding to retrieve corresponding pesticide names in the target language. Finetuning was also explored to improve the multilingual representation further. Moreover, we evaluated the use of large language models (LLMs) combined with the recent retrieval-augmented generation (RAG) framework. As a result, we found that the contextual approach, combined with fine-tuning, provided the best results, contributing to enhancing Pesticide Terminology Extraction in a multilingual scenario. © 2025, Brazilian Computing Society. All rights reserved.
5.
Merzouki, K.; Hadi, Y.; Elghazi, H.; Moudoud, H.; Houda, Z. A. El
Graph Neural Network Framework for Advanced Persistent Threat Detection in IIoT Environments Proceedings Article
In: F., El Bouanani; F., Ayoub (Ed.): Int. Conf. Adv. Commun. Technol. Netw., CommNet - Proc., Institute of Electrical and Electronics Engineers Inc., 2025, ISBN: 979-833155781-2 (ISBN), (Journal Abbreviation: Int. Conf. Adv. Commun. Technol. Netw., CommNet - Proc.).
Abstract | Links | BibTeX | Tags: Advanced persistent threat, Advanced Persistent Threats, Anomaly detection, CICAPT-IIoT2024, Deep learning, extraction, Feature extraction, Features extraction, Graph Neural Networks, Graphic methods, IIoT, Industrial infrastructure, Industrial internet of thing, Learning systems, Message passing, Network architecture, Network frameworks, Network security, Relational learning, Threat detection
@inproceedings{merzoukiGraphNeuralNetwork2025,
title = {Graph Neural Network Framework for Advanced Persistent Threat Detection in IIoT Environments},
author = {K. Merzouki and Y. Hadi and H. Elghazi and H. Moudoud and Z. A. El Houda},
editor = {El Bouanani F. and Ayoub F.},
url = {https://www.scopus.com/pages/publications/105032074689?origin=resultslist},
doi = {10.1109/CommNet68224.2025.11288886},
isbn = {979-833155781-2 (ISBN)},
year = {2025},
date = {2025-01-01},
booktitle = {Int. Conf. Adv. Commun. Technol. Netw., CommNet - Proc.},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {The increasing reliance of industrial infrastructures on the Industrial Internet of Things (IIoT) has made them more vulnerable to complex cyberattacks, especially Advanced Persistent Threats (APTs). To recognize and analyze these multi-stage incursions, we require computational models that can capture both structural and temporal connections in IIoT network architectures. This paper presents a framework based on Graph Neural Networks (GNNs) for detecting and classifying APTs in IIoT settings. We use the CICAPT-IIoT2024 dataset, which provides realistic multi-phase APT attack scenarios. The approach models system components, network communications, and process interactions as nodes and edges in a dynamic graph, allowing for relational learning and context-aware feature extraction. The GNN architecture leverages graph connectivity patterns and message-passing techniques to identify attack phases with greater accuracy and robustness. Experimental results show that this method outperforms traditional deep learning techniques and ensemble methods, particularly in early-stage anomaly detection. This paper highlights the potential of graph-based learning as an effective way to enhance IIoT infrastructure security against the changing behaviors of advanced persistent threats. © 2025 IEEE.},
note = {Journal Abbreviation: Int. Conf. Adv. Commun. Technol. Netw., CommNet - Proc.},
keywords = {Advanced persistent threat, Advanced Persistent Threats, Anomaly detection, CICAPT-IIoT2024, Deep learning, extraction, Feature extraction, Features extraction, Graph Neural Networks, Graphic methods, IIoT, Industrial infrastructure, Industrial internet of thing, Learning systems, Message passing, Network architecture, Network frameworks, Network security, Relational learning, Threat detection},
pubstate = {published},
tppubtype = {inproceedings}
}
The increasing reliance of industrial infrastructures on the Industrial Internet of Things (IIoT) has made them more vulnerable to complex cyberattacks, especially Advanced Persistent Threats (APTs). To recognize and analyze these multi-stage incursions, we require computational models that can capture both structural and temporal connections in IIoT network architectures. This paper presents a framework based on Graph Neural Networks (GNNs) for detecting and classifying APTs in IIoT settings. We use the CICAPT-IIoT2024 dataset, which provides realistic multi-phase APT attack scenarios. The approach models system components, network communications, and process interactions as nodes and edges in a dynamic graph, allowing for relational learning and context-aware feature extraction. The GNN architecture leverages graph connectivity patterns and message-passing techniques to identify attack phases with greater accuracy and robustness. Experimental results show that this method outperforms traditional deep learning techniques and ensemble methods, particularly in early-stage anomaly detection. This paper highlights the potential of graph-based learning as an effective way to enhance IIoT infrastructure security against the changing behaviors of advanced persistent threats. © 2025 IEEE.
6.
Souza, J. V.; Amamou, H.; Chen, R.; Salari, E.; Gubelmann, R.; Niklaus, C.; Serpa, T.; Lima, M. M. F.; Pinto, P. T.; Kshirsagar, S.; Davoust, A.; Handschuh, S.; Avila, A. R.
Cross-Lingual Keyword Extraction for Pesticide Terminology in Brazilian Portuguese and English Journal Article
In: Journal of the Brazilian Computer Society, vol. 31, no. 1, pp. 973–990, 2025, ISSN: 01046500 (ISSN).
Abstract | Links | BibTeX | Tags: Agriculture, BERT embedding, BERT embeddings, Cross-lingual, Embeddings, extraction, Food consumption, Keywords extraction, Labelings, Low resource languages, Multilingual extraction, Pesticides, Technical terms, Terminology, Word alignment
@article{desouzaCrossLingualKeywordExtraction2025a,
title = {Cross-Lingual Keyword Extraction for Pesticide Terminology in Brazilian Portuguese and English},
author = {J. V. Souza and H. Amamou and R. Chen and E. Salari and R. Gubelmann and C. Niklaus and T. Serpa and M. M. F. Lima and P. T. Pinto and S. Kshirsagar and A. Davoust and S. Handschuh and A. R. Avila},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-105019700300&doi=10.5753%2Fjbcs.2025.5815&partnerID=40&md5=85ee75baf4550666a307310cd04d1c83},
doi = {10.5753/jbcs.2025.5815},
issn = {01046500 (ISSN)},
year = {2025},
date = {2025-01-01},
journal = {Journal of the Brazilian Computer Society},
volume = {31},
number = {1},
pages = {973–990},
abstract = {Agriculture plays a crucial role in Brazil’s economy. As the country intensifies its activities in the sector, the use of pesticides also increases. Hence, the risks associated with pesticide-laden food consumption have become a concern for chemistry researchers. An issue affecting regulatory standardization of pesticides in Brazil is the difficulty in translating pesticide names, particularly from English. For example, the word malathion can be translated from English to Portuguese as malatiom or malatião, resulting in inconsistent labeling. This issue extends to the broader problem of translating highly technical terms between languages, in particular for low-resource languages. In this work, we investigate terminological variation in the chemistry of organophosphorus pesticides. Our goal is to study strategies for domain-specific multilingual keyword extraction. To that end, two corpora were built based on pesticide-related scientific documents in Brazilian Portuguese and English, which led to a total of 84 and 210 texts, respectively, representing the low-and high-resource languages in this study. We then assessed 6 methods for keyword extraction: Simple Maths, TF-IDF, YAKE, TextRank, MultipartiteRank, and KeyBERT. We relied on a multilingual contextual BERT embedding to retrieve corresponding pesticide names in the target language. Finetuning was also explored to improve the multilingual representation further. Moreover, we evaluated the use of large language models (LLMs) combined with the recent retrieval-augmented generation (RAG) framework. As a result, we found that the contextual approach, combined with fine-tuning, provided the best results, contributing to enhancing Pesticide Terminology Extraction in a multilingual scenario. © 2025, Brazilian Computing Society. All rights reserved.},
keywords = {Agriculture, BERT embedding, BERT embeddings, Cross-lingual, Embeddings, extraction, Food consumption, Keywords extraction, Labelings, Low resource languages, Multilingual extraction, Pesticides, Technical terms, Terminology, Word alignment},
pubstate = {published},
tppubtype = {article}
}
Agriculture plays a crucial role in Brazil’s economy. As the country intensifies its activities in the sector, the use of pesticides also increases. Hence, the risks associated with pesticide-laden food consumption have become a concern for chemistry researchers. An issue affecting regulatory standardization of pesticides in Brazil is the difficulty in translating pesticide names, particularly from English. For example, the word malathion can be translated from English to Portuguese as malatiom or malatião, resulting in inconsistent labeling. This issue extends to the broader problem of translating highly technical terms between languages, in particular for low-resource languages. In this work, we investigate terminological variation in the chemistry of organophosphorus pesticides. Our goal is to study strategies for domain-specific multilingual keyword extraction. To that end, two corpora were built based on pesticide-related scientific documents in Brazilian Portuguese and English, which led to a total of 84 and 210 texts, respectively, representing the low-and high-resource languages in this study. We then assessed 6 methods for keyword extraction: Simple Maths, TF-IDF, YAKE, TextRank, MultipartiteRank, and KeyBERT. We relied on a multilingual contextual BERT embedding to retrieve corresponding pesticide names in the target language. Finetuning was also explored to improve the multilingual representation further. Moreover, we evaluated the use of large language models (LLMs) combined with the recent retrieval-augmented generation (RAG) framework. As a result, we found that the contextual approach, combined with fine-tuning, provided the best results, contributing to enhancing Pesticide Terminology Extraction in a multilingual scenario. © 2025, Brazilian Computing Society. All rights reserved.
7.
Hebbache, L.; Amirkhani, D.; Allili, M. S.; Hammouche, N.; Lapointe, J. -F.
Leveraging Saliency in Single-Stage Multi-Label Concrete Defect Detection Using Unmanned Aerial Vehicle Imagery Journal Article
In: Remote Sensing, vol. 15, no. 5, 2023, ISSN: 20724292, (Publisher: MDPI).
Abstract | Links | BibTeX | Tags: Aerial vehicle, Aircraft detection, Antennas, Computational efficiency, Concrete defects, Deep learning, Defect detection, extraction, Feature extraction, Features extraction, Image acquisition, Image Enhancement, Multi-labels, One-stage concrete defect detection, Saliency, Single stage, Unmanned aerial vehicles (UAV), Unmanned areal vehicle imagery
@article{hebbache_leveraging_2023,
title = {Leveraging Saliency in Single-Stage Multi-Label Concrete Defect Detection Using Unmanned Aerial Vehicle Imagery},
author = {L. Hebbache and D. Amirkhani and M. S. Allili and N. Hammouche and J. -F. Lapointe},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85149966766&doi=10.3390%2frs15051218&partnerID=40&md5=7bf1cb3353270c696c07ff24dc24655d},
doi = {10.3390/rs15051218},
issn = {20724292},
year = {2023},
date = {2023-01-01},
journal = {Remote Sensing},
volume = {15},
number = {5},
publisher = {MDPI},
abstract = {Visual inspection of concrete structures using Unmanned Areal Vehicle (UAV) imagery is a challenging task due to the variability of defects’ size and appearance. This paper proposes a high-performance model for automatic and fast detection of bridge concrete defects using UAV-acquired images. Our method, coined the Saliency-based Multi-label Defect Detector (SMDD-Net), combines pyramidal feature extraction and attention through a one-stage concrete defect detection model. The attention module extracts local and global saliency features, which are scaled and integrated with the pyramidal feature extraction module of the network using the max-pooling, multiplication, and residual skip connections operations. This has the effect of enhancing the localisation of small and low-contrast defects, as well as the overall accuracy of detection in varying image acquisition ranges. Finally, a multi-label loss function detection is used to identify and localise overlapping defects. The experimental results on a standard dataset and real-world images demonstrated the performance of SMDD-Net with regard to state-of-the-art techniques. The accuracy and computational efficiency of SMDD-Net make it a suitable method for UAV-based bridge structure inspection. © 2023 by the authors.},
note = {Publisher: MDPI},
keywords = {Aerial vehicle, Aircraft detection, Antennas, Computational efficiency, Concrete defects, Deep learning, Defect detection, extraction, Feature extraction, Features extraction, Image acquisition, Image Enhancement, Multi-labels, One-stage concrete defect detection, Saliency, Single stage, Unmanned aerial vehicles (UAV), Unmanned areal vehicle imagery},
pubstate = {published},
tppubtype = {article}
}
Visual inspection of concrete structures using Unmanned Areal Vehicle (UAV) imagery is a challenging task due to the variability of defects’ size and appearance. This paper proposes a high-performance model for automatic and fast detection of bridge concrete defects using UAV-acquired images. Our method, coined the Saliency-based Multi-label Defect Detector (SMDD-Net), combines pyramidal feature extraction and attention through a one-stage concrete defect detection model. The attention module extracts local and global saliency features, which are scaled and integrated with the pyramidal feature extraction module of the network using the max-pooling, multiplication, and residual skip connections operations. This has the effect of enhancing the localisation of small and low-contrast defects, as well as the overall accuracy of detection in varying image acquisition ranges. Finally, a multi-label loss function detection is used to identify and localise overlapping defects. The experimental results on a standard dataset and real-world images demonstrated the performance of SMDD-Net with regard to state-of-the-art techniques. The accuracy and computational efficiency of SMDD-Net make it a suitable method for UAV-based bridge structure inspection. © 2023 by the authors.
8.
Cote, S. S. -P.; Paquette, G. R.; Neveu, S. -M.; Chartier, S.; Labbe, D. R.; Renaud, P.
Combining electroencephalography with plethysmography for classification of deviant sexual preferences. Proceedings Article
In: Proceedings - 9th International Workshop on Biometrics and Forensics, IWBF 2021, Institute of Electrical and Electronics Engineers Inc., 2021, ISBN: 978-172819556-8 (ISBN), (Journal Abbreviation: Proc. - Int. Workshop Biom. Forensics, IWBF).
Abstract | Links | BibTeX | Tags: Biometrics, Classification, Classification (of information), Decision trees, Deviant sexual preferences, Dimensionality reduction, Electroencephalography, Electrophysiology, extraction, Extraction method, Machine learning, Plethysmography, Proof of concept, Psychophysiological measures, Standard protocols, Variable selection and extraction
@inproceedings{cote_combining_2021,
title = {Combining electroencephalography with plethysmography for classification of deviant sexual preferences.},
author = {S. S. -P. Cote and G. R. Paquette and S. -M. Neveu and S. Chartier and D. R. Labbe and P. Renaud},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85113855965&doi=10.1109%2fIWBF50991.2021.9465078&partnerID=40&md5=b545b2a6d22e32115ac179399188960e},
doi = {10.1109/IWBF50991.2021.9465078},
isbn = {978-172819556-8 (ISBN)},
year = {2021},
date = {2021-01-01},
booktitle = {Proceedings - 9th International Workshop on Biometrics and Forensics, IWBF 2021},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {Evaluating sexual preferences is a difficult task. Past researchrelied mostly on penile plethysmography (PPG). Even though this technique is the standard protocol used in most currentforensic settings, its usage showed mixed results. One way to improve PPG is the addition of other psychophysiological measures such as electroencephalography (EEG). However, EEG generates significant amount of data that hinders classification. Machine learning (ML) is nowadays an excellent tool to identify most discriminating variables and for classification. Therefore, it is proposed to use ML selection and extraction methods for dimensionality reduction and then to classify sexual preferences. Evidence from this proof of concept shows that using EEG and PPG together leads to better classification (85.6%) than using EEG (82.2%) or PPG individually (74.4%). The Random Forest (RF) classifier combined with the Principal Component Analysis (PCA) extraction method achieves a slightly higher general performance rate. This increase in performances opens the door for using more reliable biometric measures in the assessment of deviant sexual preferences. © 2021 IEEE.},
note = {Journal Abbreviation: Proc. - Int. Workshop Biom. Forensics, IWBF},
keywords = {Biometrics, Classification, Classification (of information), Decision trees, Deviant sexual preferences, Dimensionality reduction, Electroencephalography, Electrophysiology, extraction, Extraction method, Machine learning, Plethysmography, Proof of concept, Psychophysiological measures, Standard protocols, Variable selection and extraction},
pubstate = {published},
tppubtype = {inproceedings}
}
Evaluating sexual preferences is a difficult task. Past researchrelied mostly on penile plethysmography (PPG). Even though this technique is the standard protocol used in most currentforensic settings, its usage showed mixed results. One way to improve PPG is the addition of other psychophysiological measures such as electroencephalography (EEG). However, EEG generates significant amount of data that hinders classification. Machine learning (ML) is nowadays an excellent tool to identify most discriminating variables and for classification. Therefore, it is proposed to use ML selection and extraction methods for dimensionality reduction and then to classify sexual preferences. Evidence from this proof of concept shows that using EEG and PPG together leads to better classification (85.6%) than using EEG (82.2%) or PPG individually (74.4%). The Random Forest (RF) classifier combined with the Principal Component Analysis (PCA) extraction method achieves a slightly higher general performance rate. This increase in performances opens the door for using more reliable biometric measures in the assessment of deviant sexual preferences. © 2021 IEEE.
9.
Cote, S. S. -P.; Paquette, G. R.; Neveu, S. -M.; Chartier, S.; Labbe, D. R.; Renaud, P.
Combining electroencephalography with plethysmography for classification of deviant sexual preferences. Proceedings Article
In: Proceedings - 9th International Workshop on Biometrics and Forensics, IWBF 2021, Institute of Electrical and Electronics Engineers Inc., 2021, ISBN: 978-172819556-8 (ISBN), (Journal Abbreviation: Proc. - Int. Workshop Biom. Forensics, IWBF).
Abstract | Links | BibTeX | Tags: Biometrics, Classification, Classification (of information), Decision trees, Deviant sexual preferences, Dimensionality reduction, Electroencephalography, Electrophysiology, extraction, Extraction method, Machine learning, Plethysmography, Proof of concept, Psychophysiological measures, Standard protocols, Variable selection and extraction
@inproceedings{coteCombiningElectroencephalographyPlethysmography2021,
title = {Combining electroencephalography with plethysmography for classification of deviant sexual preferences.},
author = {S. S. -P. Cote and G. R. Paquette and S. -M. Neveu and S. Chartier and D. R. Labbe and P. Renaud},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85113855965&doi=10.1109%2fIWBF50991.2021.9465078&partnerID=40&md5=b545b2a6d22e32115ac179399188960e},
doi = {10.1109/IWBF50991.2021.9465078},
isbn = {978-172819556-8 (ISBN)},
year = {2021},
date = {2021-01-01},
booktitle = {Proceedings - 9th International Workshop on Biometrics and Forensics, IWBF 2021},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {Evaluating sexual preferences is a difficult task. Past researchrelied mostly on penile plethysmography (PPG). Even though this technique is the standard protocol used in most currentforensic settings, its usage showed mixed results. One way to improve PPG is the addition of other psychophysiological measures such as electroencephalography (EEG). However, EEG generates significant amount of data that hinders classification. Machine learning (ML) is nowadays an excellent tool to identify most discriminating variables and for classification. Therefore, it is proposed to use ML selection and extraction methods for dimensionality reduction and then to classify sexual preferences. Evidence from this proof of concept shows that using EEG and PPG together leads to better classification (85.6%) than using EEG (82.2%) or PPG individually (74.4%). The Random Forest (RF) classifier combined with the Principal Component Analysis (PCA) extraction method achieves a slightly higher general performance rate. This increase in performances opens the door for using more reliable biometric measures in the assessment of deviant sexual preferences. © 2021 IEEE.},
note = {Journal Abbreviation: Proc. - Int. Workshop Biom. Forensics, IWBF},
keywords = {Biometrics, Classification, Classification (of information), Decision trees, Deviant sexual preferences, Dimensionality reduction, Electroencephalography, Electrophysiology, extraction, Extraction method, Machine learning, Plethysmography, Proof of concept, Psychophysiological measures, Standard protocols, Variable selection and extraction},
pubstate = {published},
tppubtype = {inproceedings}
}
Evaluating sexual preferences is a difficult task. Past researchrelied mostly on penile plethysmography (PPG). Even though this technique is the standard protocol used in most currentforensic settings, its usage showed mixed results. One way to improve PPG is the addition of other psychophysiological measures such as electroencephalography (EEG). However, EEG generates significant amount of data that hinders classification. Machine learning (ML) is nowadays an excellent tool to identify most discriminating variables and for classification. Therefore, it is proposed to use ML selection and extraction methods for dimensionality reduction and then to classify sexual preferences. Evidence from this proof of concept shows that using EEG and PPG together leads to better classification (85.6%) than using EEG (82.2%) or PPG individually (74.4%). The Random Forest (RF) classifier combined with the Principal Component Analysis (PCA) extraction method achieves a slightly higher general performance rate. This increase in performances opens the door for using more reliable biometric measures in the assessment of deviant sexual preferences. © 2021 IEEE.
10.
Charbonneau, I.; Robinson, K.; Blais, C.; Fiset, D.
Implicit race attitudes modulate visual information extraction for trustworthiness judgments Journal Article
In: PLoS ONE, vol. 15, no. 9 September, 2020, ISSN: 19326203, (Publisher: Public Library of Science).
Abstract | Links | BibTeX | Tags: adult, African American, African Americans, article, Attitude, Caucasian, decision making, Ethics, European Continental Ancestry Group, extraction, eyelash, Facial Expression, facies, female, human, Humans, Judgment, male, perception, physiology, psychology, Racism, Social Perception, Stereotyping, visual information, wrinkle, Young Adult
@article{charbonneau_implicit_2020,
title = {Implicit race attitudes modulate visual information extraction for trustworthiness judgments},
author = {I. Charbonneau and K. Robinson and C. Blais and D. Fiset},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85091622106&doi=10.1371%2fjournal.pone.0239305&partnerID=40&md5=18ca2332affc9cb41d17afc8c450b0b4},
doi = {10.1371/journal.pone.0239305},
issn = {19326203},
year = {2020},
date = {2020-01-01},
journal = {PLoS ONE},
volume = {15},
number = {9 September},
publisher = {Public Library of Science},
abstract = {Black people are still considered to be one of the most stigmatized groups and have to face multiple prejudices that undermine their well-being. Assumptions and beliefs about other racial groups are quite pervasive and have been shown to impact basic social tasks such as face processing. For example, individuals with high racial prejudice conceptualize other-race faces as less trustworthy and more criminal. However, it is unknown if implicit racial bias could modulate even low-level perceptual mechanisms such as spatial frequency (SF) extraction when judging the level of trustworthiness of other-race faces. The present study showed that although similar facial features are used to judge the trustworthiness of White and Black faces, own-race faces are processed in lower SF (i.e. coarse information such as the contour of the face and blurred shapes as opposed to high SF representing fine-grained information such as eyelashes or fine wrinkles). This pattern was modulated by implicit race biases: higher implicit biases are associated with a significantly higher reliance on low SF with White than with Black faces. Copyright: © 2020 Charbonneau et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.},
note = {Publisher: Public Library of Science},
keywords = {adult, African American, African Americans, article, Attitude, Caucasian, decision making, Ethics, European Continental Ancestry Group, extraction, eyelash, Facial Expression, facies, female, human, Humans, Judgment, male, perception, physiology, psychology, Racism, Social Perception, Stereotyping, visual information, wrinkle, Young Adult},
pubstate = {published},
tppubtype = {article}
}
Black people are still considered to be one of the most stigmatized groups and have to face multiple prejudices that undermine their well-being. Assumptions and beliefs about other racial groups are quite pervasive and have been shown to impact basic social tasks such as face processing. For example, individuals with high racial prejudice conceptualize other-race faces as less trustworthy and more criminal. However, it is unknown if implicit racial bias could modulate even low-level perceptual mechanisms such as spatial frequency (SF) extraction when judging the level of trustworthiness of other-race faces. The present study showed that although similar facial features are used to judge the trustworthiness of White and Black faces, own-race faces are processed in lower SF (i.e. coarse information such as the contour of the face and blurred shapes as opposed to high SF representing fine-grained information such as eyelashes or fine wrinkles). This pattern was modulated by implicit race biases: higher implicit biases are associated with a significantly higher reliance on low SF with White than with Black faces. Copyright: © 2020 Charbonneau et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.