Jay (JieBing) Yu, PhD’s Post

🍵 Some weekend reading from the team at RelationalAI…. We want to share a paper, “Leapfrog Triejoin: A Simple, Worst-Case Optimal Join Algorithm,” by our own Todd Veldhuizen, honored with the ICDT 2024 Test of Time award. According to the 2024 Awards Committee, “the ICDT 2014 paper has greatly influenced theoretical as well as practical database research and is, in fact, one of the great gems of ICDT". 💎 A true classic. Have a read. https://hubs.li/Q02Q3Q9v0

Comparing the already existing segmentation methods for masking diseased spots on grapevine leaves: “Segmentation Methods Evaluation on Grapevine Leaf Diseases“ by Szilard Molnar, Levente Tamas. Proceedings of the 18th Conference on Computer Science and Intelligence Systems(Thematic Tracks Short Papers), appeared in: Maria Ganzha, L. Maciaszek, Marcin Paprzycki, Dominik Ślęzak (eds); ACSIS, Vol. 35, pages 1081–1085(2023). #vinediseasedetection #segmentationmethods #leafdiseases Open Access: https://lnkd.in/dHKQpknF PTI - Polskie Towarzystwo Informatyczne QED SOFTWARE

🎉 Excited to share that our paper: "SimSCOOD: Systematic Analysis of Out-of-Distribution Generalization in Fine-tuned Source Code Models" has been accepted in NAACL 2024 Findings! In this paper, we contribute the first systematic approach that simulates various OOD scenarios along different dimensions of source code data properties and study the fine-tuned model behaviors in such scenarios. Paper: https://lnkd.in/eExnDgc7 Many thanks to my co-authors Ning Yu, Cristian-Alexandru Staicu, and Mario Fritz

Exciting news! We just published a groundbreaking blog post on a novel Clustering Method for Maximizing Decoding Information within graph-based models. The post delves into the innovative CMDI, which leverages two-dimensional structural information theory to address the uncertainty associated with random walk access between nodes and the embedded structural information in the data. Empirical evaluations demonstrate that CMDI outperforms classical baseline methods, showcasing superior decoding information ratio and heightened efficiency, particularly when considering prior knowledge. Dive into the details and explore the future of graph-based clustering analysis at https://bit.ly/3xdciOG.

Alhmdulliah ❤️ "Automatic Bengali Number Plate Detection and Authentication using YOLO-V4 and YOLO-V5" this research paper Published in: 2023 26th International Conference on Computer and Information Technology (ICCIT) . That will help to solve the problem, first detecting the license plate and extracting the license plate region from the image and then passing this image to the number plate class detection model to extract the information from the license plate image.

Semantic web in libraries

I’m very happy to share that our* paper History repeats itself: A Baseline for Temporal Knowledge Graph Forecasting was accepted at #IJCAI2024. 🎉 *Julia Gastinger, Christian Meilicke, Federico Errica, Timo Sztyler, Anett Schuelke, Heiner Stuckenschmidt In the paper we propose a baseline for Temporal Knowledge Graph Forecasting, leveraging the principle of recurrency. Surprisingly, this simple approach beats state of the art methods on 3 out of 5 commonly used datasets.   Please find the paper here: https://lnkd.in/e7eqMQ9c

Did you know that a simple #temporal #knowledge #graph #baseline can beat state of the art models on three out of five benchmark datasets? If not, check out our accepted paper at IJCAI 2024! Shout out to Julia for leading this project!!

Learn about the Open Science Framework (OSF), a free, open-source tool designed to help researchers manage projects, collaborate seamlessly, and share data openly. Register for webinar: https://lnkd.in/e_wqTv3w In this webinar, you'll: ► Explore use cases demonstrating how OSF supports open science ► Learn how OSF addresses common challenges faced by researchers ► Receive a guided tour of key workflows and features

My first article has been published in Future Internet MDPI, with José Carlos Metrôlho, PhD and Fernando Reinaldo Ribeiro. The article introduces Karmem, a solution for overcoming challenges in WebAssembly, enabling data transfer and module interaction through a new interface description language and an optimized application binary interface. You can read the full article here: https://lnkd.in/dcAcFem5