Closely linked to the topic of ML is data science, which is of particular interest for approaches in machine learning and As a consequence of the increasing influence of thus also reflected in multiple AI curricula. Corresponding machine learning on our lives, everyone needs methods are also used to gain knowledge in a wide variety competencies to understand corresponding phenomena, of scientific disciplines. Data analysis and artificial intelligence but also to get involved in shaping our are often referred to as the fourth pillar of science world and making informed decisions regarding (Riedel et al., 2008; Tolle et al., 2011). This is becoming the influences on our society. Therefore, in K-increasingly relevant for K-12 education as well, as this shift 12 education, students need to learn about core in the scientific disciplines is also reflected in corresponding ideas and principles of machine learning.

Adaptive learning is an area of educational technology that consists in delivering personalized learning experiences to address the unique needs of each learner. An important subfield of adaptive learning is learning path personalization: it aims at designing systems that recommend sequences of educational activities to maximize students' learning outcomes. Many machine learning approaches have already demonstrated significant results in a variety of contexts related to learning path personalization. However, most of them were designed for very specific settings and are not very reusable. This is accentuated by the fact that they often rely on non-scalable models, which are unable to integrate new elements after being trained on a specific set of educational resources. In this paper, we introduce a flexible and scalable approach towards the problem of learning path personalization, which we formalize as a reinforcement learning problem. Our model is a sequential recommender system based on a graph neural network, which we evaluate on a population of simulated learners. Our results demonstrate that it can learn to make good recommendations in the small-data regime.

In this paper, we propose a novel architecture called Composition Attention Grammars (CAGs) that recursively compose subtrees into a single vector representation with a composition function, and selectively attend to previous structural information with a self-attention mechanism. We investigate whether these components -- the composition function and the self-attention mechanism -- can both induce human-like syntactic generalization. Specifically, we train language models (LMs) with and without these two components with the model sizes carefully controlled, and evaluate their syntactic generalization performance against six test circuits on the SyntaxGym benchmark. The results demonstrated that the composition function and the self-attention mechanism both play an important role to make LMs more human-like, and closer inspection of linguistic phenomenon implied that the composition function allowed syntactic features, but not semantic features, to percolate into subtree representations.

Large-scale pre-trained language models have shown outstanding performance in a variety of NLP tasks. However, they are also known to be significantly brittle against specifically crafted adversarial examples, leading to increasing interest in probing the adversarial robustness of NLP systems. We introduce RSMI, a novel two-stage framework that combines randomized smoothing (RS) with masked inference (MI) to improve the adversarial robustness of NLP systems. RS transforms a classifier into a smoothed classifier to obtain robust representations, whereas MI forces a model to exploit the surrounding context of a masked token in an input sequence. RSMI improves adversarial robustness by 2 to 3 times over existing state-of-the-art methods on benchmark datasets. We also perform in-depth qualitative analysis to validate the effectiveness of the different stages of RSMI and probe the impact of its components through extensive ablations. By empirically proving the stability of RSMI, we put it forward as a practical method to robustly train large-scale NLP models. Our code and datasets are available at https://github.com/Han8931/rsmi_nlp

The Quran is the holy scripture of Islam, and its recitation is an important aspect of the religion. Recognizing the recitation of the Holy Quran automatically is a challenging task due to its unique rules that are not applied in normal speaking speeches. A lot of research has been done in this domain, but previous works have detected recitation errors as a classification task or used traditional automatic speech recognition (ASR). In this paper, we proposed a novel end-to-end deep learning model for recognizing the recitation of the Holy Quran. The proposed model is a CNN-Bidirectional GRU encoder that uses CTC as an objective function, and a character-based decoder which is a beam search decoder. Moreover, all previous works were done on small private datasets consisting of short verses and a few chapters of the Holy Quran. As a result of using private datasets, no comparisons were done. To overcome this issue, we used a public dataset that has recently been published (Ar-DAD) and contains about 37 chapters that were recited by 30 reciters, with different recitation speeds and different types of pronunciation rules. The proposed model performance was evaluated using the most common evaluation metrics in speech recognition, word error rate (WER), and character error rate (CER). The results were 8.34% WER and 2.42% CER. We hope this research will be a baseline for comparisons with future research on this public new dataset (Ar-DAD).

Rep. Michael McCaul, R-Texas, said it is essential for the United States to protect its intellectual property and loosen its reliance on China's supply chain to win the race for quantum computing, artificial intelligence (AI) and semiconductor chip supremacy. You know, we had the Russians and we won that race. We have to win this one," McCaul told Fox News Digital at the Milken Global Conference. McCaul, the House Foreign Affairs Committee Chair, said that whoever gets to quantum first is going to rule the world. Additionally, China has been very clear that their 100-year goal is to enact complete "military and economic domination of the world," a mission that McCaul asserted the U.S. cannot allow to happen. U.S. investors have pushed billions into China's AI sector, a February report from CSET showed. China is currently investing heavily in quantum computing, AI and advanced weapons systems. As such, McCaul stressed the importance of U.S. corporations working to protect their intellectual property. "We have to stop exporting our technology to China that they can put in things like the hypersonic missile, for instance, or the spy ballon, for that matter, had American parts in it, component parts," McCaul said. Despite the national security concerns, McCaul noted that the U.S. can work with China's extensive market as long the country understands the critical supply chains intertwining the two nations. One of these critical supply chains involves semiconductors. McCaul first introduced the CHIPS for America Act in 2020, and it was signed into law in August 2022. The Act provides $280 billion in new funding to bolster domestic manufacturing and research for semiconductors domestically. McCaul said the Act is part of a broader effort to pull semiconductor reliance out of Taiwan and South Korea. In fact, TSMC, the world's largest semiconductor foundry, is located on the island of Taiwan. "You know, I introduced the CHIPS bill to try to move some of that out of country.

There's a chance that ChatGPT knows personal details about you--and if it doesn't, it might just make something up. As OpenAI's generative text chatbot has boomed in popularity over the past six months, the risks of the system being trained on data vacuumed up from the web have become clearer. Data regulators around the world are investigating issues with how OpenAI gathered the data it uses to train its large language models, the accuracy of answers it provides about people, and other legal concerns about the use of its generative text systems. Europe's data regulators have joined forces to look at OpenAI after Italy temporarily banned ChatGPT from the country. And Canada is also investigating the technology's potential privacy risks.

Large AI Models (LAIMs), of which large language models are the most prominent recent example, showcase some impressive performance. However they have been empirically found to pose serious security issues. This paper systematizes our knowledge about the fundamental impossibility of building arbitrarily accurate and secure machine learning models. More precisely, we identify key challenging features of many of today's machine learning settings. Namely, high accuracy seems to require memorizing large training datasets, which are often user-generated and highly heterogeneous, with both sensitive information and fake users. We then survey statistical lower bounds that, we argue, constitute a compelling case against the possibility of designing high-accuracy LAIMs with strong security guarantees.

Goal-directed dialogue systems aim to proactively reach a pre-determined target through multi-turn conversations. The key to achieving this task lies in planning dialogue paths that smoothly and coherently direct conversations towards the target. However, this is a challenging and under-explored task. In this work, we propose a coherent dialogue planning approach that uses a stochastic process to model the temporal dynamics of dialogue paths. We define a latent space that captures the coherence of goal-directed behavior using a Brownian bridge process, which allows us to incorporate user feedback flexibly in dialogue planning. Based on the derived latent trajectories, we generate dialogue paths explicitly using pre-trained language models. We finally employ these paths as natural language prompts to guide dialogue generation. Our experiments show that our approach generates more coherent utterances and achieves the goal with a higher success rate.

The NLP community recently saw the release of a new large open-access multilingual language model, BLOOM (BigScience et al., 2022) covering 46 languages. We focus on BLOOM's multilingual ability by evaluating its machine translation performance across several datasets (WMT, Flores-101 and DiaBLa) and language pairs (high- and low-resourced). Our results show that 0-shot performance suffers from overgeneration and generating in the wrong language, but this is greatly improved in the few-shot setting, with very good results for a number of language pairs. We study several aspects including prompt design, model sizes, cross-lingual transfer and the use of discursive context.