At Microsoft, we believe artificial intelligence (AI) is the defining technology of our time. We have been on the forefront of cutting-edge research in AI and integrating these powerful, innovative AI technologies into our products and services to help customers do more. Microsoft AI, powered by Azure, provides billions of intelligent experiences every day in Windows, Xbox, Microsoft 365, Teams, Azure AI, Power Platform, Dynamics 365 and Microsoft Defender. Our AI tools and technologies are designed to benefit everyone at every level in every organization. They are used in workplaces, home offices, academic institutions, research labs and manufacturing facilities around the world, and they are helping everyone from scientists and salespeople to farmers, software developers and security practitioners.

The use of AI/ML capabilities for improving HCI/UX work and delivering better UX in solutions is becoming a trend (Abbas et al., 2022; Wu et al., 2019; Nikiforova et al., 2021) and creates many new opportunities for HCI/UX professionals (Holmquist, 2017; Yang et al., 2020). Some even speculate "AI/ML is the new UX" (Yang et al., 2018). Researchers proposed that AI can perform as an assistant, collaborator, researcher, or facilitator (Bertão & Joo, 2021; Main & Grierson, 2020). AI technology will change the role of designers in the design process and generate an opportunity for creative collaboration between AI and designers (McCormack et al., 2020). Also, companies are moving fast to adopt AI for improving customer experience (CX).

In recent years, graph representation learning has achieved remarkable success while suffering from low-quality data problems. As a mature technology to improve data quality in computer vision, data augmentation has also attracted increasing attention in graph domain. For promoting the development of this emerging research direction, in this survey, we comprehensively review and summarize the existing graph data augmentation (GDAug) techniques. Specifically, we first summarize a variety of feasible taxonomies, and then classify existing GDAug studies based on fine-grained graph elements. Furthermore, for each type of GDAug technique, we formalize the general definition, discuss the technical details, and give schematic illustration. In addition, we also summarize common performance metrics and specific design metrics for constructing a GDAug evaluation system. Finally, we summarize the applications of GDAug from both data and model levels, as well as future directions. Latest advances in GDAug are summarized in a GitHub repository: https://github.com/jjzhou012/GDAug-Survey.

Future AI applications require performance, reliability and privacy that the existing, cloud-dependant system architectures cannot provide. In this article, we study orchestration in the device-edge-cloud continuum, and focus on edge AI for resource orchestration. We claim that to support the constantly growing requirements of intelligent applications in the device-edge-cloud computing continuum, resource orchestration needs to embrace edge AI and emphasize local autonomy and intelligence. To justify the claim, we provide a general definition for continuum orchestration, and look at how current and emerging orchestration paradigms are suitable for the computing continuum. We describe certain major emerging research themes that may affect future orchestration, and provide an early vision of an orchestration paradigm that embraces those research themes. Finally, we survey current key edge AI methods and look at how they may contribute into fulfilling the vision of future continuum orchestration.

Microstructure reconstruction is an important and emerging field of research and an essential foundation to improving inverse computational materials engineering (ICME). Much of the recent progress in the field is made based on generative adversarial networks (GANs). Although excellent results have been achieved throughout a variety of materials, challenges remain regarding the interpretability of the model's latent space as well as the applicability to extremely small data sets. The present work addresses these issues by introducing DA-VEGAN, a model with two central innovations. First, a $\beta$-variational autoencoder is incorporated into a hybrid GAN architecture that allows to penalize strong nonlinearities in the latent space by an additional parameter, $\beta$. Secondly, a custom differentiable data augmentation scheme is developed specifically for this architecture. The differentiability allows the model to learn from extremely small data sets without mode collapse or deteriorated sample quality. An extensive validation on a variety of structures demonstrates the potential of the method and future directions of investigation are discussed.

Modern, state-of-the-art Convolutional Neural Networks (CNNs) in computer vision have millions of parameters. Thus, explaining the complex decisions of such networks to humans is challenging. A technical approach to reduce CNN complexity is network pruning, where less important parameters are deleted. The work presented in this paper investigates whether this technical complexity reduction also helps with perceived explainability. To do so, we conducted a pre-study and two human-grounded experiments, assessing the effects of different pruning ratios on CNN explainability. Overall, we evaluated four different compression rates (i.e., CPR 2, 4, 8, and 32) with 37 500 tasks on Mechanical Turk. Results indicate that lower compression rates have a positive influence on explainability, while higher compression rates show negative effects. Furthermore, we were able to identify sweet spots that increase both the perceived explainability and the model's performance.

To mitigate the effects of undesired biases in models, several approaches propose to pre-process the input dataset to reduce the risks of discrimination by preventing the inference of sensitive attributes. Unfortunately, most of these pre-processing methods lead to the generation a new distribution that is very different from the original one, thus often leading to unrealistic data. As a side effect, this new data distribution implies that existing models need to be re-trained to be able to make accurate predictions. To address this issue, we propose a novel pre-processing method, that we coin as fair mapping, based on the transformation of the distribution of protected groups onto a chosen target one, with additional privacy constraints whose objective is to prevent the inference of sensitive attributes. More precisely, we leverage on the recent works of the Wasserstein GAN and AttGAN frameworks to achieve the optimal transport of data points coupled with a discriminator enforcing the protection against attribute inference. Our proposed approach, preserves the interpretability of data and can be used without defining exactly the sensitive groups. In addition, our approach can be specialized to model existing state-of-the-art approaches, thus proposing a unifying view on these methods. Finally, several experiments on real and synthetic datasets demonstrate that our approach is able to hide the sensitive attributes, while limiting the distortion of the data and improving the fairness on subsequent data analysis tasks.

Distributed quantum information processing protocols such as quantum entanglement distillation and quantum state discrimination rely on local operations and classical communications (LOCC). Existing LOCC-based protocols typically assume the availability of ideal, noiseless, communication channels. In this paper, we study the case in which classical communication takes place over noisy channels, and we propose to address the design of LOCC protocols in this setting via the use of quantum machine learning tools. We specifically focus on the important tasks of quantum entanglement distillation and quantum state discrimination, and implement local processing through parameterized quantum circuits (PQCs) that are optimized to maximize the average fidelity and average success probability in the respective tasks, while accounting for communication errors. The introduced approach, Noise Aware-LOCCNet (NA-LOCCNet), is shown to have significant advantages over existing protocols designed for noiseless communications.

Retrobiosynthesis provides an effective and sustainable approach to producing functional molecules. The past few decades have witnessed a rapid expansion of biosynthetic approaches. With the recent advances in data-driven sciences, machine learning (ML) is enriching the retrobiosynthesis design toolbox and being applied to each step of the synthesis design workflow, including retrosynthesis planning, enzyme identification and engineering, and pathway optimization. The ability to learn from existing knowledge, recognize complex patterns and generalize to the unknown has made ML a promising solution to biological problems. In this Review, we summarize the recent progress in the development of ML models for assisting with molecular synthesis. We highlight the key advantages of ML-based biosynthesis design methods and discuss the challenges and outlook for the further development of ML-based approaches. Retrobiosynthesis aims to create novel biosynthetic pathways for the beneficial production of molecules of interest. This Review outlines how machine learning can help to advance retrobiosynthesis by improving retrosynthesis planning, enzyme identification and selection, and the engineering of enzymes and pathways.

This survey reviews works in which language models (LMs) are augmented with reasoning skills and the ability to use tools. The former is defined as decomposing a potentially complex task into simpler subtasks while the latter consists in calling external modules such as a code interpreter. LMs can leverage these augmentations separately or in combination via heuristics, or learn to do so from demonstrations. While adhering to a standard missing tokens prediction objective, such augmented LMs can use various, possibly non-parametric external modules to expand their context processing ability, thus departing from the pure language modeling paradigm. We therefore refer to them as Augmented Language Models (ALMs). The missing token objective allows ALMs to learn to reason, use tools, and even act, while still performing standard natural language tasks and even outperforming most regular LMs on several benchmarks. In this work, after reviewing current advance in ALMs, we conclude that this new research direction has the potential to address common limitations of traditional LMs such as interpretability, consistency, and scalability issues.