About
Technology & AI leader with a passion for product and digital transformation. Bridge between tech, product and business. Academic background in physics and information studies. Over 15 years of programming and building digital products.
Activity
-
Ik werk inmiddels al geruime tijd als onderzoeker in AI en Data Science, en ik zie dagelijks hoe waardevol onze oplossingen zijn. Onze technologieën…
Ik werk inmiddels al geruime tijd als onderzoeker in AI en Data Science, en ik zie dagelijks hoe waardevol onze oplossingen zijn. Onze technologieën…
Liked by Floris Hermsen
-
🌟 Inside the GENIUS Lab: Working on the Future of AI 🌟 We’re excited to share more about the GENIUS Lab, a collaboration between Kickstart AI, TU…
🌟 Inside the GENIUS Lab: Working on the Future of AI 🌟 We’re excited to share more about the GENIUS Lab, a collaboration between Kickstart AI, TU…
Liked by Floris Hermsen
Experience
Education
Publications
-
An End-to-End Pipeline for Bibliography Extraction from Scientific Articles
IJCNLP-AACL 2023
We introduce a comprehensive end-to-end pipeline designed to extract complete bibliography section from English scientific articles in digital-born PDF format and further split them into individual citations. At the heart of our pipeline lies the utilization of Language-independent Layout Transformer (LiLT), a multimodal model that combines text and layout features to enhance the accuracy and robustness of bibliography extraction. By considering both text and visual structure, LiLT…
We introduce a comprehensive end-to-end pipeline designed to extract complete bibliography section from English scientific articles in digital-born PDF format and further split them into individual citations. At the heart of our pipeline lies the utilization of Language-independent Layout Transformer (LiLT), a multimodal model that combines text and layout features to enhance the accuracy and robustness of bibliography extraction. By considering both text and visual structure, LiLT significantly improves the identification of bibliographic sections within scientific articles. To split the extracted full bibliography into individual citations, we employ a custom fine-tuned version of SciBERT, a Transformer-based model that excels at handling complex formatting variations common in scholarly bibliography.
-
End-to-End Learning from Complex Multigraphs with Latent-Graph Convolutional Networks
We study the problem of end-to-end learning from complex multigraphs with potentially very large numbers of edges between two vertices, each edge labeled with rich information. Examples range from communication networks to flights between airports or financial transaction graphs. We propose Latent-Graph Convolutional Networks (L-GCNs), which propagate information from these complex edges to a latent adjacency tensor, after which further downstream tasks can be performed, such as node…
We study the problem of end-to-end learning from complex multigraphs with potentially very large numbers of edges between two vertices, each edge labeled with rich information. Examples range from communication networks to flights between airports or financial transaction graphs. We propose Latent-Graph Convolutional Networks (L-GCNs), which propagate information from these complex edges to a latent adjacency tensor, after which further downstream tasks can be performed, such as node classification. We evaluate the performance of several variations of the model on two synthetic datasets simulating fraud in financial transaction networks, ensuring the model must make use of edge labels in order to achieve good classification performance. We find that allowing for nonlinear interactions on a per-neighbor basis boosts performance significantly, while showing promising results in an inductive setting. Finally, we demonstrate the use of L-GCNs on real-world data in the form of an urban transportation network.
Languages
-
Dutch
Native or bilingual proficiency
-
English
Full professional proficiency
-
German
Professional working proficiency
-
French
Elementary proficiency
More activity by Floris
𝗛𝗼𝘁 𝘁𝗮𝗸𝗲: 𝗮𝗿𝗲 𝗟𝗮𝗿𝗴𝗲 𝗟𝗮𝗻𝗴𝘂𝗮𝗴𝗲 𝗠𝗼𝗱𝗲𝗹𝘀 𝙖𝙘𝙩𝙪𝙖𝙡𝙡𝙮 𝗶𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝘁? Many fellow AI experts I talk to argue…
Shared by Floris Hermsen
We are building the world’s fastest, holistic search engine for drug development and scientific professionals, devised for them and designed with…
Liked by Floris Hermsen
🌳 It’s not easy being green 🐸 Earlier this year, the European Parliament gave a final green light to a new law that will ban misleading…
Liked by Floris Hermsen
🌟 Let's embrace innovative approach to elevate Data Reliability Diving into the era of generative AI, we uncover its potential to revolutionize…
Liked by Floris Hermsen
Really good to have been a part of the #TLCGlobal #CDO Series earlier this week, hosted by Rela8 Group Ltd. A great opportunity to delve into the…
Shared by Floris Hermsen
Interesting article on the restrictions by the #euaiact for #opensourceai models. While open-source AI models are exempted when used for scientific…
Liked by Floris Hermsen
Whenever I get the opportunity to meet people who have taken the leap into entrepreneurial journey I am excited to hear their experience, but even…
Liked by Floris Hermsen
🚀 Stoked to share that I've joined Hugging Face! Throughout the years I've been impressed by how Hugging Face has been empowering the AI community…
Liked by Floris Hermsen
People also viewed
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore MoreOthers named Floris Hermsen
2 others named Floris Hermsen are on LinkedIn
See others named Floris Hermsen