Slide
Centre Interdisciplinaire
de Recherche et d’Innovation
en Cybersécurité et Société
de Recherche et d’Innovation
en Cybersécurité et Société
1.
Gagnon, S.
"Use of AI to Fight Disinformation and Corruption" Proceedings Article
In: Am. Conf. Inf. Syst., AMCIS, pp. 4853–4854, Association for Information Systems, 2025, ISBN: 979-833132774-3 (ISBN).
Abstract | Links | BibTeX | Tags: Action research, Audio videos, Copyrights, Crime, Due diligence, Information leakage, Information use, Institutional change, Integrated strategy, Knowledge graph, Knowledge graphs, Open source intelligence, Open systems, Research advances
@inproceedings{gagnon_use_2025,
title = {"Use of AI to Fight Disinformation and Corruption"},
author = {S. Gagnon},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-105025353954&partnerID=40&md5=fc3ca73e0de02b06fd218e03588f3e22},
isbn = {979-833132774-3 (ISBN)},
year = {2025},
date = {2025-01-01},
booktitle = {Am. Conf. Inf. Syst., AMCIS},
volume = {7},
pages = {4853–4854},
publisher = {Association for Information Systems},
abstract = {Disinformation (false information disseminated with the intent to cause harm) takes many forms, including fake news, audio-video impersonations, data falsification, and more. It is linked to allegations of corruption and fraud against elected officials, which, once disambiguated (whether true or false), will in all cases require immediate action to preserve democracy. An integrated strategy to combat disinformation and corruption is needed, requiring constant vigilance, due diligence, and open-source intelligence (OSINT) collaboration. Our PDS will explore how AI, especially LLMs and Knowledge Graphs, can serve this purpose and automate the necessary means to combat hybrid threats to democracy. The PDS will cover the following issues and latest research advances, allowing also interactive roundtables to ensure participants can contribute insight to the debate: 1. Presentation: fighting jointly against disinformation and corruption 2. Roundtable: what innovations are most effective in combating hybrid threats? 3. Demo: ai and knowledge graphs to disambiguate allegations against elected officials 4. Roundtable: what institutional changes are required to fight hybrid threats? 5. Conclusion: action research helping parliamentary institutions develop AI capabilities. Copyright © 2025 by Association for Information Systems (AIS). All rights reserved.},
keywords = {Action research, Audio videos, Copyrights, Crime, Due diligence, Information leakage, Information use, Institutional change, Integrated strategy, Knowledge graph, Knowledge graphs, Open source intelligence, Open systems, Research advances},
pubstate = {published},
tppubtype = {inproceedings}
}
Disinformation (false information disseminated with the intent to cause harm) takes many forms, including fake news, audio-video impersonations, data falsification, and more. It is linked to allegations of corruption and fraud against elected officials, which, once disambiguated (whether true or false), will in all cases require immediate action to preserve democracy. An integrated strategy to combat disinformation and corruption is needed, requiring constant vigilance, due diligence, and open-source intelligence (OSINT) collaboration. Our PDS will explore how AI, especially LLMs and Knowledge Graphs, can serve this purpose and automate the necessary means to combat hybrid threats to democracy. The PDS will cover the following issues and latest research advances, allowing also interactive roundtables to ensure participants can contribute insight to the debate: 1. Presentation: fighting jointly against disinformation and corruption 2. Roundtable: what innovations are most effective in combating hybrid threats? 3. Demo: ai and knowledge graphs to disambiguate allegations against elected officials 4. Roundtable: what institutional changes are required to fight hybrid threats? 5. Conclusion: action research helping parliamentary institutions develop AI capabilities. Copyright © 2025 by Association for Information Systems (AIS). All rights reserved.
2.
Gagnon, S.
"Use of AI to Fight Disinformation and Corruption" Proceedings Article
In: Am. Conf. Inf. Syst., AMCIS, pp. 4853–4854, Association for Information Systems, 2025, ISBN: 979-833132774-3 (ISBN), (Journal Abbreviation: Am. Conf. Inf. Syst., AMCIS).
Abstract | Links | BibTeX | Tags: Action research, Audio videos, Copyrights, Crime, Due diligence, Information leakage, Information use, Institutional change, Integrated strategy, Knowledge graph, Knowledge graphs, Open source intelligence, Open systems, Research advances
@inproceedings{gagnonUseAIFight2025,
title = {"Use of AI to Fight Disinformation and Corruption"},
author = {S. Gagnon},
url = {https://www.scopus.com/pages/publications/105025353954?origin=resultslist},
isbn = {979-833132774-3 (ISBN)},
year = {2025},
date = {2025-01-01},
booktitle = {Am. Conf. Inf. Syst., AMCIS},
volume = {7},
pages = {4853–4854},
publisher = {Association for Information Systems},
abstract = {Disinformation (false information disseminated with the intent to cause harm) takes many forms, including fake news, audio-video impersonations, data falsification, and more. It is linked to allegations of corruption and fraud against elected officials, which, once disambiguated (whether true or false), will in all cases require immediate action to preserve democracy. An integrated strategy to combat disinformation and corruption is needed, requiring constant vigilance, due diligence, and open-source intelligence (OSINT) collaboration. Our PDS will explore how AI, especially LLMs and Knowledge Graphs, can serve this purpose and automate the necessary means to combat hybrid threats to democracy. The PDS will cover the following issues and latest research advances, allowing also interactive roundtables to ensure participants can contribute insight to the debate: 1. Presentation: fighting jointly against disinformation and corruption 2. Roundtable: what innovations are most effective in combating hybrid threats? 3. Demo: ai and knowledge graphs to disambiguate allegations against elected officials 4. Roundtable: what institutional changes are required to fight hybrid threats? 5. Conclusion: action research helping parliamentary institutions develop AI capabilities. Copyright © 2025 by Association for Information Systems (AIS). All rights reserved.},
note = {Journal Abbreviation: Am. Conf. Inf. Syst., AMCIS},
keywords = {Action research, Audio videos, Copyrights, Crime, Due diligence, Information leakage, Information use, Institutional change, Integrated strategy, Knowledge graph, Knowledge graphs, Open source intelligence, Open systems, Research advances},
pubstate = {published},
tppubtype = {inproceedings}
}
Disinformation (false information disseminated with the intent to cause harm) takes many forms, including fake news, audio-video impersonations, data falsification, and more. It is linked to allegations of corruption and fraud against elected officials, which, once disambiguated (whether true or false), will in all cases require immediate action to preserve democracy. An integrated strategy to combat disinformation and corruption is needed, requiring constant vigilance, due diligence, and open-source intelligence (OSINT) collaboration. Our PDS will explore how AI, especially LLMs and Knowledge Graphs, can serve this purpose and automate the necessary means to combat hybrid threats to democracy. The PDS will cover the following issues and latest research advances, allowing also interactive roundtables to ensure participants can contribute insight to the debate: 1. Presentation: fighting jointly against disinformation and corruption 2. Roundtable: what innovations are most effective in combating hybrid threats? 3. Demo: ai and knowledge graphs to disambiguate allegations against elected officials 4. Roundtable: what institutional changes are required to fight hybrid threats? 5. Conclusion: action research helping parliamentary institutions develop AI capabilities. Copyright © 2025 by Association for Information Systems (AIS). All rights reserved.
3.
Amamou, H.; Gagnon, S.; Davoust, A.; Avila, A. R.
Towards Robust Retrieval-Augmented Generation Based on Knowledge Graph: A Comparative Analysis Proceedings Article
In: Conf. Proc. IEEE Int. Conf. Syst. Man Cybern., pp. 80–85, Institute of Electrical and Electronics Engineers Inc., 2025, ISBN: 1062922X (ISSN); 979-833153358-8 (ISBN), (Journal Abbreviation: Conf. Proc. IEEE Int. Conf. Syst. Man Cybern.).
Abstract | Links | BibTeX | Tags: Benchmarking, Comparative analyzes, External sources, Generation systems, Information integration, Information retrieval, Knowledge graph, Knowledge graphs, Language model, Model response, Noise robustness, Pre-training, Prior-knowledge
@inproceedings{amamouRobustRetrievalAugmentedGeneration2025,
title = {Towards Robust Retrieval-Augmented Generation Based on Knowledge Graph: A Comparative Analysis},
author = {H. Amamou and S. Gagnon and A. Davoust and A. R. Avila},
url = {https://www.scopus.com/pages/publications/105033145604?origin=resultslist},
doi = {10.1109/SMC58881.2025.11343466},
isbn = {1062922X (ISSN); 979-833153358-8 (ISBN)},
year = {2025},
date = {2025-01-01},
booktitle = {Conf. Proc. IEEE Int. Conf. Syst. Man Cybern.},
pages = {80–85},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {Retrieval-Augmented Generation (RAG) was first introduced to enhance the capabilities of Large Language Models (LLMs) beyond their encoded-prior knowledge. This is achieved by providing LLMs with an external source of knowledge, which helps to reduce factual hallucinations and enables the access to new information, typically not available during their pretraining phase. Despite its benefits, there is an increasing concern with the impact of inconsistent retrieved information towards LLMs' responses. Hence, the Retrieval-Augmented Generation Benchmark (RGB) was introduced as a new testbed for RAG evaluation, meant to assess the robustness of LLMs towards inconsistency in the retrieved information. In this work, we use the RGB corpus to evaluate LLMs in four scenarios: (1) noise robustness; (2) information integration; (3) negative rejection; and (4) counterfactual robustness. We perform a comparative analysis between the RAG baseline defined by the RGB and variations of GraphRAG, which is a RAG system based on a Knowledge Graph (KG) and developed to retrieve relevant information from large documents. We tested GraphRAG under three customization to improve its robustness. Our approach demonstrates improvements compared to the RGB baseline, providing insights on how to design more reliable RAG systems, tailored for real-world scenarios. © 2025 IEEE.},
note = {Journal Abbreviation: Conf. Proc. IEEE Int. Conf. Syst. Man Cybern.},
keywords = {Benchmarking, Comparative analyzes, External sources, Generation systems, Information integration, Information retrieval, Knowledge graph, Knowledge graphs, Language model, Model response, Noise robustness, Pre-training, Prior-knowledge},
pubstate = {published},
tppubtype = {inproceedings}
}
Retrieval-Augmented Generation (RAG) was first introduced to enhance the capabilities of Large Language Models (LLMs) beyond their encoded-prior knowledge. This is achieved by providing LLMs with an external source of knowledge, which helps to reduce factual hallucinations and enables the access to new information, typically not available during their pretraining phase. Despite its benefits, there is an increasing concern with the impact of inconsistent retrieved information towards LLMs' responses. Hence, the Retrieval-Augmented Generation Benchmark (RGB) was introduced as a new testbed for RAG evaluation, meant to assess the robustness of LLMs towards inconsistency in the retrieved information. In this work, we use the RGB corpus to evaluate LLMs in four scenarios: (1) noise robustness; (2) information integration; (3) negative rejection; and (4) counterfactual robustness. We perform a comparative analysis between the RAG baseline defined by the RGB and variations of GraphRAG, which is a RAG system based on a Knowledge Graph (KG) and developed to retrieve relevant information from large documents. We tested GraphRAG under three customization to improve its robustness. Our approach demonstrates improvements compared to the RGB baseline, providing insights on how to design more reliable RAG systems, tailored for real-world scenarios. © 2025 IEEE.
4.
Gagnon, S.
"Use of AI to Fight Disinformation and Corruption" Proceedings Article
In: Am. Conf. Inf. Syst., AMCIS, pp. 4853–4854, Association for Information Systems, 2025, ISBN: 979-833132774-3 (ISBN).
Abstract | Links | BibTeX | Tags: Action research, Audio videos, Copyrights, Crime, Due diligence, Information leakage, Information use, Institutional change, Integrated strategy, Knowledge graph, Knowledge graphs, Open source intelligence, Open systems, Research advances
@inproceedings{gagnonUseAIFight2025a,
title = {"Use of AI to Fight Disinformation and Corruption"},
author = {S. Gagnon},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-105025353954&partnerID=40&md5=fc3ca73e0de02b06fd218e03588f3e22},
isbn = {979-833132774-3 (ISBN)},
year = {2025},
date = {2025-01-01},
booktitle = {Am. Conf. Inf. Syst., AMCIS},
volume = {7},
pages = {4853–4854},
publisher = {Association for Information Systems},
abstract = {Disinformation (false information disseminated with the intent to cause harm) takes many forms, including fake news, audio-video impersonations, data falsification, and more. It is linked to allegations of corruption and fraud against elected officials, which, once disambiguated (whether true or false), will in all cases require immediate action to preserve democracy. An integrated strategy to combat disinformation and corruption is needed, requiring constant vigilance, due diligence, and open-source intelligence (OSINT) collaboration. Our PDS will explore how AI, especially LLMs and Knowledge Graphs, can serve this purpose and automate the necessary means to combat hybrid threats to democracy. The PDS will cover the following issues and latest research advances, allowing also interactive roundtables to ensure participants can contribute insight to the debate: 1. Presentation: fighting jointly against disinformation and corruption 2. Roundtable: what innovations are most effective in combating hybrid threats? 3. Demo: ai and knowledge graphs to disambiguate allegations against elected officials 4. Roundtable: what institutional changes are required to fight hybrid threats? 5. Conclusion: action research helping parliamentary institutions develop AI capabilities. Copyright © 2025 by Association for Information Systems (AIS). All rights reserved.},
keywords = {Action research, Audio videos, Copyrights, Crime, Due diligence, Information leakage, Information use, Institutional change, Integrated strategy, Knowledge graph, Knowledge graphs, Open source intelligence, Open systems, Research advances},
pubstate = {published},
tppubtype = {inproceedings}
}
Disinformation (false information disseminated with the intent to cause harm) takes many forms, including fake news, audio-video impersonations, data falsification, and more. It is linked to allegations of corruption and fraud against elected officials, which, once disambiguated (whether true or false), will in all cases require immediate action to preserve democracy. An integrated strategy to combat disinformation and corruption is needed, requiring constant vigilance, due diligence, and open-source intelligence (OSINT) collaboration. Our PDS will explore how AI, especially LLMs and Knowledge Graphs, can serve this purpose and automate the necessary means to combat hybrid threats to democracy. The PDS will cover the following issues and latest research advances, allowing also interactive roundtables to ensure participants can contribute insight to the debate: 1. Presentation: fighting jointly against disinformation and corruption 2. Roundtable: what innovations are most effective in combating hybrid threats? 3. Demo: ai and knowledge graphs to disambiguate allegations against elected officials 4. Roundtable: what institutional changes are required to fight hybrid threats? 5. Conclusion: action research helping parliamentary institutions develop AI capabilities. Copyright © 2025 by Association for Information Systems (AIS). All rights reserved.
5.
Azzi, S.; Gagnon, S.
Ontology-Driven Parliamentary Analytics: Analysing Political Debates on COVID-19 Impact in Canada Journal Article
In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 14149 LNCS, pp. 89–102, 2023, ISSN: 03029743, (ISBN: 9783031398407).
Abstract | Links | BibTeX | Tags: COVID-19, End-users, Knowledge graph, Knowledge graphs, Ontology, Ontology graphs, Ontology's, Parliamentary debate, Political debates, Question Answering, Semantic content, Semantic ontology, Semantics, Solid basis
@article{azzi_ontology-driven_2023,
title = {Ontology-Driven Parliamentary Analytics: Analysing Political Debates on COVID-19 Impact in Canada},
author = {S. Azzi and S. Gagnon},
editor = {Asemi A. Francesconi E. Ko A.},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85172088757&doi=10.1007%2f978-3-031-39841-4_7&partnerID=40&md5=5dff3b672c36c66070042a35eed048e0},
doi = {10.1007/978-3-031-39841-4_7},
issn = {03029743},
year = {2023},
date = {2023-01-01},
journal = {Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)},
volume = {14149 LNCS},
pages = {89–102},
publisher = {Springer Science and Business Media Deutschland GmbH},
abstract = {Parliamentary debates are usually published in Parliament’s websites to allows citizens to be informed on the latest national debates, proposals and decisions. To enhance citizen experience and engagement, functionalities such as debates annotation and question answering are necessary. Annotating text requires semantic content and ontologies are known for their ability to describe a common vocabulary for a domain and can be a solid base for annotation and question answering. We report on an ongoing study to enhance parliamentary analytics using an ontology and knowledge graph to sharpen annotations and facilitate their query by end-users. As a salient case, a sample of debates are collected on the COVID-19 impact in Canada, as its complexity shows the relevance of using advanced knowledge representation techniques. We focused on the development of a new “Impact of COVID-19 in Canada Ontology” (ICCO) that provides contextualized semantic information on impact in numerous policy areas, as this ontology is entirely built from Canadian parliamentary debates. It has been evaluated and validated by experts. Our conclusion underscores the importance of integrating ontology-driven parliamentary analytics within the broader context of digital transformation in legislative institutions, and the need for new platforms supporting free and open Digital Humanities. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.},
note = {ISBN: 9783031398407},
keywords = {COVID-19, End-users, Knowledge graph, Knowledge graphs, Ontology, Ontology graphs, Ontology's, Parliamentary debate, Political debates, Question Answering, Semantic content, Semantic ontology, Semantics, Solid basis},
pubstate = {published},
tppubtype = {article}
}
Parliamentary debates are usually published in Parliament’s websites to allows citizens to be informed on the latest national debates, proposals and decisions. To enhance citizen experience and engagement, functionalities such as debates annotation and question answering are necessary. Annotating text requires semantic content and ontologies are known for their ability to describe a common vocabulary for a domain and can be a solid base for annotation and question answering. We report on an ongoing study to enhance parliamentary analytics using an ontology and knowledge graph to sharpen annotations and facilitate their query by end-users. As a salient case, a sample of debates are collected on the COVID-19 impact in Canada, as its complexity shows the relevance of using advanced knowledge representation techniques. We focused on the development of a new “Impact of COVID-19 in Canada Ontology” (ICCO) that provides contextualized semantic information on impact in numerous policy areas, as this ontology is entirely built from Canadian parliamentary debates. It has been evaluated and validated by experts. Our conclusion underscores the importance of integrating ontology-driven parliamentary analytics within the broader context of digital transformation in legislative institutions, and the need for new platforms supporting free and open Digital Humanities. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
6.
Gagnon, S.; Azzi, S.
Semantic Annotation of Parliamentary Debates and Legislative Intelligence Enhancing Citizen Experience Journal Article
In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 13429 LNCS, pp. 63–76, 2022, ISSN: 03029743, (ISBN: 9783031126727).
Abstract | Links | BibTeX | Tags: Analytic tools, Core functionality, Digital solutions, Knowledge graph, Language processing, Laws and legislation, Legislative intelligence, Natural language processing systems, Natural languages, Parliamentary debate, Parliamentary proceedings, Semantic annotations, Semantic-analytics, Semantics
@article{gagnon_semantic_2022,
title = {Semantic Annotation of Parliamentary Debates and Legislative Intelligence Enhancing Citizen Experience},
author = {S. Gagnon and S. Azzi},
editor = {Kotsis G. Francesconi E. Ko A.},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85135782631&doi=10.1007%2f978-3-031-12673-4_5&partnerID=40&md5=0765bacc7d38f77896bd9adf402268b9},
doi = {10.1007/978-3-031-12673-4_5},
issn = {03029743},
year = {2022},
date = {2022-01-01},
journal = {Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)},
volume = {13429 LNCS},
pages = {63–76},
publisher = {Springer Science and Business Media Deutschland GmbH},
abstract = {The concept of “Legislative Intelligence” (LegisIntel) refers to Artificial Intelligence (AI) and semantic analytics tools implemented in parliaments to enhance citizen experience in monitoring complex interrelations among various contents of parliamentary proceedings. The integration of a suite of digital solutions can build upon the core functionality of Semantic Annotation of Parliamentary Debates. Using well-established Natural Language Processing (NLP) technologies, linked to ontologies and Knowledge Graphs (KG), it can help identify the concepts and entities throughout texts, and index sentences and summaries as per a citizen’s customized knowledge base. These annotations can then be leveraged to recommend relevant text excerpts end-users could build upon, within teams if they chose to do so, and possibly compose and customize legislative critiques and recommendations thoroughly tested for coherence, accuracy, and evidence. The present study proposes an international open-source initiative among parliaments to ensure the launch and viability of a suite of LegisIntel solutions. It reports on the completed initial phase of this initiative, aiming to prepare discussions in launching an international consultation among peers. The goals of this phase are to document the core functionality of LegisIntel solutions and formulate a proposed architecture that may serve to generate ideas from various developer communities. The Action Design Research (ADR) methodology is used in this process, with results focused on system artefacts such as an interface mockup, a functional design, and a model of infrastructure components. The conclusion addresses risks and outlines the next steps of this initiative. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.},
note = {ISBN: 9783031126727},
keywords = {Analytic tools, Core functionality, Digital solutions, Knowledge graph, Language processing, Laws and legislation, Legislative intelligence, Natural language processing systems, Natural languages, Parliamentary debate, Parliamentary proceedings, Semantic annotations, Semantic-analytics, Semantics},
pubstate = {published},
tppubtype = {article}
}
The concept of “Legislative Intelligence” (LegisIntel) refers to Artificial Intelligence (AI) and semantic analytics tools implemented in parliaments to enhance citizen experience in monitoring complex interrelations among various contents of parliamentary proceedings. The integration of a suite of digital solutions can build upon the core functionality of Semantic Annotation of Parliamentary Debates. Using well-established Natural Language Processing (NLP) technologies, linked to ontologies and Knowledge Graphs (KG), it can help identify the concepts and entities throughout texts, and index sentences and summaries as per a citizen’s customized knowledge base. These annotations can then be leveraged to recommend relevant text excerpts end-users could build upon, within teams if they chose to do so, and possibly compose and customize legislative critiques and recommendations thoroughly tested for coherence, accuracy, and evidence. The present study proposes an international open-source initiative among parliaments to ensure the launch and viability of a suite of LegisIntel solutions. It reports on the completed initial phase of this initiative, aiming to prepare discussions in launching an international consultation among peers. The goals of this phase are to document the core functionality of LegisIntel solutions and formulate a proposed architecture that may serve to generate ideas from various developer communities. The Action Design Research (ADR) methodology is used in this process, with results focused on system artefacts such as an interface mockup, a functional design, and a model of infrastructure components. The conclusion addresses risks and outlines the next steps of this initiative. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.