Researchers address the challenge of limited Arabic medical dialogue data by generating 80,000 synthetic question-answer pairs using ChatGPT-4o and Gemini 2.5 Pro, expanding an initial dataset of 20,000 records. They fine-tuned five LLMs, including Mistral-7B and AraGPT2, and evaluated performance using BERTScore and expert review. Results showed that training with ChatGPT-4o-generated data led to higher F1-scores and fewer hallucinations across models. Why it matters: This demonstrates the potential of synthetic data augmentation to improve domain-specific Arabic language models, particularly for low-resource medical NLP applications.
This paper introduces neural Bayes estimators for censored peaks-over-threshold models, enhancing computational efficiency in spatial extremal dependence modeling. The method uses data augmentation to encode censoring information in the neural network input, challenging traditional likelihood-based approaches. The estimators were applied to assess extreme particulate matter concentrations over Saudi Arabia, demonstrating efficacy in high-dimensional models. Why it matters: The research offers a computationally efficient alternative for environmental modeling and risk assessment in the region.
KAUST Ph.D. student Matthias Müller won the Best Paper/Presentation Award at the 2nd International Workshop on Computer Vision for Unmanned Aerial Vehicles for his paper on teaching UAVs to navigate a racecourse autonomously. The paper, "Teaching UAVs to Race: End to End Regression of Agile Controls in Simulation," details research on training a deep neural network to predict UAV controls from raw image data. The research uses imitation learning with data augmentation to allow for correction of navigation mistakes, outperforming state-of-the-art methods. Why it matters: This award recognizes KAUST's contributions to computer vision and autonomous drone navigation, important areas for future applications in logistics, surveillance, and environmental monitoring in the region.
Thamar Solorio from the University of Houston will discuss machine learning approaches for spontaneous human language processing. The talk will cover adapting multilingual transformers to code-switching data and using data augmentation for domain adaptation in sequence labeling tasks. Solorio will also provide an overview of other research projects at the RiTUAL lab, focusing on the scarcity of labeled data. Why it matters: This presentation addresses key challenges in Arabic NLP related to data scarcity, which is a persistent obstacle in developing effective AI applications for the region.
A CMU professor and MBZUAI affiliated faculty presented research on how LLMs store and use knowledge learned during pre-training. The study used a synthetic biography dataset to show that LLMs may not effectively use memorized knowledge at inference time, even with zero training loss. Data augmentation during pre-training can force the model to store knowledge in specific token embeddings. Why it matters: The research highlights limitations in LLM knowledge manipulation and extraction, with implications for improving model architectures and training strategies for more effective knowledge utilization in Arabic LLMs.
Injy Hamed from NYU Abu Dhabi's CAMeL Lab presented work on Egyptian Arabic-English code-switching for ASR and MT. She discussed the ArzEn-ST speech translation corpus and compared end-to-end and hybrid systems for ASR. For MT, she presented data augmentation and word segmentation techniques to handle data scarcity, also addressing ASR evaluation challenges in code-switching. Why it matters: Research into code-switching is crucial for building NLP systems capable of processing real-world language use in the Arab world.