In task-oriented conversation systems, natural language generation systems that generate sentences with specific information related to conversation flow are useful. Our study focuses on language generation by considering various information representing the meaning of utterances as multiple conditions of generation. NLG from meaning representations, the conditions for sentence meaning, generally goes through two steps: sentence planning and surface realization. However, we propose a simple one-stage framework to generate utterances directly from MR (Meaning Representation). Our model is based on GPT2 and generates utterances with flat conditions on slot and value pairs, which does not need to determine the structure of the sentence. We evaluate several systems in the E2E dataset with 6 automatic metrics. Our system is a simple method, but it demonstrates comparable performance to previous systems in automated metrics. In addition, using only 10\% of the data set without any other techniques, our model achieves comparable performance, and shows the possibility of performing zero-shot generation and expanding to other datasets.
We propose a Distributional Approach to address Controlled Text Generation from pre-trained Language Models (LMs). This view permits to define, in a single formal framework, "pointwise" and "distributional" constraints over the target LM -- to our knowledge, this is the first approach with such generality -- while minimizing KL divergence with the initial LM distribution. The optimal target distribution is then uniquely determined as an explicit EBM (Energy-Based Model) representation. From that optimal representation we then train the target controlled autoregressive LM through an adaptive distributional variant of Policy Gradient. We conduct a first set of experiments over pointwise constraints showing the advantages of our approach over a set of baselines, in terms of obtaining a controlled LM balancing constraint satisfaction with divergence from the initial LM (GPT-2). We then perform experiments over distributional constraints, a unique feature of our approach, demonstrating its potential as a remedy to the problem of Bias in Language Models. Through an ablation study we show the effectiveness of our adaptive technique for obtaining faster convergence.
Learning to control the structure of sentences is a challenging problem in text generation. Existing work either relies on simple deterministic approaches or RL-based hard structures. We explore the use of structured variational autoencoders to infer latent templates for sentence generation using a soft, continuous relaxation in order to utilize reparameterization for training. Specifically, we propose a Gumbel-CRF, a continuous relaxation of the CRF sampling algorithm using a relaxed Forward-Filtering Backward-Sampling (FFBS) approach. As a reparameterized gradient estimator, the Gumbel-CRF gives more stable gradients than score-function based estimators. As a structured inference network, we show that it learns interpretable templates during training, which allows us to control the decoder during testing. We demonstrate the effectiveness of our methods with experiments on data-to-text generation and unsupervised paraphrase generation.
High-dimensional representations for words, text, images, knowledge graphs and other structured data are commonly used in different paradigms of machine learning and data mining. These representations have different degrees of interpretability, with efficient distributed representations coming at the cost of the loss of feature to dimension mapping. This implies that there is obfuscation in the way concepts are captured in these embedding spaces. Its effects are seen in many representations and tasks, one particularly problematic one being in language representations where the societal biases, learned from underlying data, are captured and occluded in unknown dimensions and subspaces. As a result, invalid associations (such as different races and their association with a polar notion of good versus bad) are made and propagated by the representations, leading to unfair outcomes in different tasks where they are used. This work addresses some of these problems pertaining to the transparency and interpretability of such representations. A primary focus is the detection, quantification, and mitigation of socially biased associations in language representation.
Popular task-oriented dialog data sets such as MultiWOZ (Budzianowski et al. 2018) are created by providing crowd-sourced workers a goal instruction, expressed in natural language, that describes the task to be accomplished. Crowd-sourced workers play the role of a user and an agent to generate dialogs to accomplish tasks involving booking restaurant tables, making train reservations, calling a taxi etc. However, creating large crowd-sourced datasets can be time consuming and expensive. To reduce the cost associated with generating such dialog datasets, recent work has explored methods to automatically create larger datasets from small samples.In this paper, we present a data creation strategy that uses the pre-trained language model, GPT2 (Radford et al. 2018), to simulate the interaction between crowd-sourced workers by creating a user bot and an agent bot. We train the simulators using a smaller percentage of actual crowd-generated conversations and their corresponding goal instructions. We demonstrate that by using the simulated data, we achieve significant improvements in both low-resource setting as well as in over-all task performance. To the best of our knowledge we are the first to present a model for generating entire conversations by simulating the crowd-sourced data collection process
Ice & Fries, a Viking-themed fast casual concept in Iceland, relies on drink-making robots, a bionic dog and 3D technology to provide an epic self-service dining experience. Ice & Fries features two cocktail shaking and dancing robots that can make more than 150 drinks per hour.Editor's note: An earlier version of this article ran on Fast Casual, a Kiosk Marketplace sister publication.Most fast casual restaurants are known for their atmosphere as well as food, but that's an understatement for Ice & Fries, a Viking-themed restaurant and bar in Iceland.Operating mostly as a self-service concept, Ice & Fries opened in March boasting two cocktail-shaking and dancing robots -- Tipsy Floki and Ragnar -- who can make more than 150 drinks per hour.The menus are projected onto glass windows at the entrance, and as customers step forward, light projection way-finders on the floor illuminate the experience that lies ahead. The seating area has an iceberg-inspired design featuring sound auditoriums where guests sit directly under domes to hear curated playlists.
Over time, different models have emerged to help us to solve DM problems. In particular, multi-person multi-criteria decision making (MpMcDM) models consider the evaluations of multiple experts to solve a decision situation analyzing all possible solution alternatives according to several criteria . Computational DM process, as the human DM one, requires of useful, complete and insightful information for making the most adequate decision according to the input information. The input of DM models is usually a set of evaluations from the experts. They wish to express their evaluations in natural language, but raw text is not directly processed by DM models. Accordingly, several approaches are followed for asking and elaborating a computational representation of the evaluations, namely: (1) using a numerical representation of the evaluations  and (2) using a predefined set of linguistic terms . These approaches for asking evaluations constrain the evaluative expressiveness of the experts, because they have to adapt their evaluation to the numerical or linguistic evaluation alternatives. We claim that experts in a DM problem have to express their evaluations in natural language, and the DM model has to be able to process and computationally represent them. Natural language processing (NLP) is the artificial intelligence area that combines linguistic and computational language backgrounds for understanding and generating human language [16, 28].
The company behind the world's first AI-powered robot kitchen assistant has announced its debut funding round in the UK in what could be a pivotal step in its quest to get the concept established with restaurant chains here. Miso Robotics – the US creator of the Flippy robot – is aiming to raise £24m via Crowdcube to support its expansion into Europe. The company has previously raised more than $17m (£13m) in funding rounds in the US, following a valuation of over £64m in 2019. Flippy, which cooks burgers, fries and chicken, can learn from its surroundings and acquire new skills and is already deployed in the US market at CaliBurger restaurants and iconic venues such as the Dodger Stadium in Los Angeles through Levy Restaurants, part of Compass Group. This week, Miso Robotics announced that US fast food chain White Castle will deploy Flippy in order to modernise its operations. The fundraising comes at a time when QSRs are having to work even harder to build resilient operations that offer safer working environments as they reopen following the Covid-19 pandemic.
With this paper, we aim to put an issue on the agenda of AI ethics that in our view is overlooked in the current discourse. The current discussions are dominated by topics suchas trustworthiness and bias, whereas the issue we like to focuson is counter to the debate on trustworthiness. We fear that the overuse of currently dominant AI systems that are driven by short-term objectives and optimized for avoiding error leads to a society that loses its diversity and flexibility needed for true progress. We couch our concerns in the discourse around the term anti-fragility and show with some examples what threats current methods used for decision making pose for society.
In many applications labeled data is not readily available, and needs to be collected via painstaking human supervision. We propose a rule-exemplar method for collecting human supervision to combine the efficiency of rules with the quality of instance labels. The supervision is coupled such that it is both natural for humans and synergistic for learning. We propose a training algorithm that jointly denoises rules via latent coverage variables, and trains the model through a soft implication loss over the coverage and label variables. The denoised rules and trained model are used jointly for inference. Empirical evaluation on five different tasks shows that (1) our algorithm is more accurate than several existing methods of learning from a mix of clean and noisy supervision, and (2) the coupled rule-exemplar supervision is effective in denoising rules. With the ever-increasing reach of machine learning, a common hurdle to new adoptions is the lack of labeled data and the painstaking process involved in collecting human supervision. Over the years, several strategies have evolved.