2024-07-05, 09:00–12:00 (Europe/Paris), Workshop room 1
Natural Language Processing (NLP) is a vital, well-established tool across sectors like finance and risk management. Within cybersecurity, NLP is crucial for extracting and analyzing information, data annotation, developing detection protocols, and malware classification. Today some of its key roles include extracting indicators of compromise from unstructured data and automating threat intelligence reporting.
Contrary to popular belief, the field is open to everyone; with some programming skills and a dash of creativity, anyone can start and get some results pretty quickly.
This workshop is an introduction for total beginners, where no prior knowledge in natural language processing or machine learning is required.
We break down NLP concepts and go back to the core of it, reducing complexity of NLP/LLM projects to the understanding essentials of how language models work and how to structure your code.
Key takeaways:
- lay the foundations for a good understanding of natural language processing
- develop an intuition of what can be done with NLP in cybersecurity
- equip participants with tools to start coding right away: development environment, frameworks, and some available datasets
- teach how to approach an NLP problem
- pointers on resources and code from the opensource community
You will learn how to code a full natural language processing pipeline from scratch. We will cover two examples of tasks, one for text classification and the other for text generation.
You will learn how to choose and use a model, as well as evaluate its performance.
This workshop is 100% hands-on.
The ecosystem being largely opensource, we aim also to introduce participants on how to contribute to ML. Why not by submitting work merging the fields of NLP and cybersecurity?
With a background in criminology and software engineering, Pauline harbors a strong passion for linguistics, for which she was trained at the University of Sorbonne. She leads Cubessa, where her work focuses on the intersection of AI, human cognition, and cybersecurity. With her unique linguistics-based perspective, she brings innovative insights into the development of AI systems and their role in cybersecurity. More than just a technical contributor, Pauline is a fervent advocate for AI education and actively involved in open-source projects. Her commitment extends to conducting hardware security training sessions and participating in the MISP community. Additionally, she co-founded the DEFCON group in Paris.
Previously working as a Senior Threat Intelligence analyst, she continues to dedicate her work to the field of human-centered security, an approach that emphasizes the human aspect in cybersecurity and focuses on understanding how human behavior interacts with and affects security systems.