Designers of online deliberative platforms aim to counter the degrading quality of online debates and eliminate online discrimination based on class, race or gender. Support technologies such as machine learning and natural language processing open avenues for widening the circle of people involved in deliberation, moving from small groups to ``crowd'' scale. Some design features of large-scale online discussion systems allow larger numbers of people to discuss shared problems, enhance critical thinking, and formulate solutions. However, scaling up deliberation is challenging. We review the transdisciplinary literature on the design of digital mass-deliberation platforms and examine the commonly featured design aspects (e.g., argumentation support, automated facilitation, and gamification). We find that the literature is heavily focused on developing technical fixes for scaling up deliberation, with a heavy western influence on design and test users skew young and highly educated. Contrastingly, there is a distinct lack of discussion on the nature of the design process, the inclusion of stakeholders and issues relating to inclusion, which may unwittingly perpetuate bias. Another tendency of deliberation platforms is to nudge participants to desired forms of argumentation, and simplifying definitions of good and bad arguments to fit algorithmic purposes. Few studies bridge disciplines between deliberative theory, design and engineering. As a result, scaling up deliberation will likely advance in separate systemic siloes. We make design and process recommendations to correct this course and suggest avenues for future research.

Alongside huge volumes of research on deep learning models in NLP in the recent years, there has been also much work on benchmark datasets needed to track modeling progress. Question answering and reading comprehension have been particularly prolific in this regard, with over 80 new datasets appearing in the past two years. This study is the largest survey of the field to date. We provide an overview of the various formats and domains of the current resources, highlighting the current lacunae for future work. We further discuss the current classifications of ``reasoning types" in question answering and propose a new taxonomy. We also discuss the implications of over-focusing on English, and survey the current monolingual resources for other languages and multilingual resources. The study is aimed at both practitioners looking for pointers to the wealth of existing data, and at researchers working on new resources.

Unsupervised domain adaptation (UDA) for semantic segmentation aims to adapt a segmentation model trained on the labeled source domain to the unlabeled target domain. Existing methods try to learn domain invariant features while suffering from large domain gaps that make it difficult to correctly align discrepant features, especially in the initial training phase. To address this issue, we propose a novel Dual Soft-Paste (DSP) method in this paper. Specifically, DSP selects some classes from a source domain image using a long-tail class first sampling strategy and softly pastes the corresponding image patch on both the source and target training images with a fusion weight. Technically, we adopt the mean teacher framework for domain adaptation, where the pasted source and target images go through the student network while the original target image goes through the teacher network. Output-level alignment is carried out by aligning the probability maps of the target fused image from both networks using a weighted cross-entropy loss. In addition, feature-level alignment is carried out by aligning the feature maps of the source and target images from student network using a weighted maximum mean discrepancy loss. DSP facilitates the model learning domain-invariant features from the intermediate domains, leading to faster convergence and better performance. Experiments on two challenging benchmarks demonstrate the superiority of DSP over state-of-the-art methods. Code is available at \url{https://github.com/GaoLii/DSP}.

Backpacks to track bees, bushfire modelling and sensors to detect broken water pipes are some of the technologies being developed in Australia as part of an artificial intelligence boom. Digital Economy Minister Jane Hume says artificial intelligence has the capacity to improve the lives of all Australians. Senator Hume will tell a Committee for Economic Development of Australia event on Tuesday the government had two roles to play in terms of AI: an enabler and a standards setter, especially in terms of ethics. Earlier this year the government announced a $1.2 billion digital economic strategy which included an additional $124 million commitment to the AI initiatives. The CSIRO estimates AI technology will contribute $22 trillion into the global economy by 2030.

Eyenuk, Inc., a global artificial intelligence (AI) medical technology and services company and the leader in real-world applications for AI Eye Screening, announced that its EyeArt AI system for diabetic eye testing has been chosen for deployment in 4 hospitals in Binh Dinh Province, Vietnam. The project is funded by The Fred Hollows Foundation. "We are excited to implement the EyeArt AI system to expand our capabilities in detection and treatment of diabetic retinopathy. It will help us reach our goal to protect the vision of approximately six million people with diabetes living in Vietnam," said Pham Quoc Anh, Vietnam Country Manager for The Fred Hollows Foundation. "This important project will help us continue the work first started by Professor Fred Hollows 29 years ago, fulfilling his vision to bring equitable eye health for all."

Conversational agents, or chatbots, providing question-answer assistance on smart devices, have proliferated in recent years and are poised to transform online customer services of corporate sectors.1,6 Implemented through dialogue management systems, chatbots converse through voice-based and textual dialogue, and harness natural language processing and artificial intelligence to recognize requests, provide responses, and predict user behavior.5,28 Market analysts concur on current adoption trends and the magnitude of growth and impact of chatbots anticipated in the next five years. According to a report by Grand View Research, for instance, already 45% of users prefer chatbots as the primary point of communications for customer service enquiries, translating into a global'chatbot' market of $1.23 billion by 2025, at a compounded annual growth rate (CAGR) of 24.3%.9 The strategy for conducting conversations using chatbots requires an efficient resolution of two key aspects. First, user queries or automatically perceived needs through user interactions have to be interpreted and mapped into categories, or user intents. This is based on historical processing of queries and needs, and the use of intent classification techniques.12 Second, conversations must be constructed for specific intents using frame-based dialogue management2 and neural response generation techniques.15 In frame-based dialogue management, the chatbot needs to converse with the user to have a fully filled frame (for example, flight information) in which all slot values are provided by the user (for example, airline carrier, departure time, departure location, and arrival location). The dialogue flow is constructed through an ordered sequence of frames.

This paper presents a novel approach to automated playtesting for the prediction of human player behavior and experience. It has previously been demonstrated that Deep Reinforcement Learning (DRL) game-playing agents can predict both game difficulty and player engagement, operationalized as average pass and churn rates. We improve this approach by enhancing DRL with Monte Carlo Tree Search (MCTS). We also motivate an enhanced selection strategy for predictor features, based on the observation that an AI agent's best-case performance can yield stronger correlations with human data than the agent's average performance. Both additions consistently improve the prediction accuracy, and the DRL-enhanced MCTS outperforms both DRL and vanilla MCTS in the hardest levels. We conclude that player modelling via automated playtesting can benefit from combining DRL and MCTS. Moreover, it can be worthwhile to investigate a subset of repeated best AI agent runs, if AI gameplay does not yield good predictions on average.

Meta-learning has emerged as a trending technique to tackle few-shot text classification and achieved state-of-the-art performance. However, existing solutions heavily rely on the exploitation of lexical features and their distributional signatures on training data, while neglecting to strengthen the model's ability to adapt to new tasks. In this paper, we propose a novel meta-learning framework integrated with an adversarial domain adaptation network, aiming to improve the adaptive ability of the model and generate high-quality text embedding for new classes. Extensive experiments are conducted on four benchmark datasets and our method demonstrates clear superiority over the state-of-the-art models in all the datasets. In particular, the accuracy of 1-shot and 5-shot classification on the dataset of 20 Newsgroups is boosted from 52.1% to 59.6%, and from 68.3% to 77.8%, respectively.

When humans solve complex problems, they rarely come up with a decision right-away. Instead, they start with an intuitive decision, reflect upon it, spot mistakes, resolve contradictions and jump between different hypotheses. Thus, they create a sequence of ideas and follow a train of thought that ultimately reaches a conclusive decision. Contrary to this, today's neural classification models are mostly trained to map an input to one single and fixed output. In this paper, we investigate how we can give models the opportunity of a second, third and $k$-th thought. We take inspiration from Hegel's dialectics and propose a method that turns an existing classifier's class prediction (such as the image class forest) into a sequence of predictions (such as forest $\rightarrow$ tree $\rightarrow$ mushroom). Concretely, we propose a correction module that is trained to estimate the model's correctness as well as an iterative prediction update based on the prediction's gradient. Our approach results in a dynamic system over class probability distributions $\unicode{x2014}$ the thought flow. We evaluate our method on diverse datasets and tasks from computer vision and natural language processing. We observe surprisingly complex but intuitive behavior and demonstrate that our method (i) can correct misclassifications, (ii) strengthens model performance, (iii) is robust to high levels of adversarial attacks, (iv) can increase accuracy up to 4% in a label-distribution-shift setting and (iv) provides a tool for model interpretability that uncovers model knowledge which otherwise remains invisible in a single distribution prediction.

Despite the success of deep neural networks (DNNs) in enabling on-device voice assistants, increasing evidence of bias and discrimination in machine learning is raising the urgency of investigating the fairness of these systems. Speaker verification is a form of biometric identification that gives access to voice assistants. Due to a lack of fairness metrics and evaluation frameworks that are appropriate for testing the fairness of speaker verification components, little is known about how model performance varies across subgroups, and what factors influence performance variation. To tackle this emerging challenge, we design and develop SVEva Fair, an accessible, actionable and model-agnostic framework for evaluating the fairness of speaker verification components. The framework provides evaluation measures and visualisations to interrogate model performance across speaker subgroups and compare fairness between models. We demonstrate SVEva Fair in a case study with end-to-end DNNs trained on the VoxCeleb datasets to reveal potential bias in existing embedded speech recognition systems based on the demographic attributes of speakers. Our evaluation shows that publicly accessible benchmark models are not fair and consistently produce worse predictions for some nationalities, and for female speakers of most nationalities. To pave the way for fair and reliable embedded speaker verification, SVEva Fair has been implemented as an open-source python library and can be integrated into the embedded ML development pipeline to facilitate developers and researchers in troubleshooting unreliable speaker verification performance, and selecting high impact approaches for mitigating fairness challenges