Assessing the quality of natural language generation systems through human annotation is very expensive. Additionally, human annotation campaigns are time-consuming and include non-reusable human labour. In practice, researchers rely on automatic metrics as a proxy of quality. In the last decade, many string-based metrics (e.g., BLEU) have been introduced. However, such metrics usually rely on exact matches and thus, do not robustly handle synonyms. In this paper, we introduce InfoLM a family of untrained metrics that can be viewed as a string-based metric that addresses the aforementioned flaws thanks to a pre-trained masked language model. This family of metrics also makes use of information measures allowing the adaptation of InfoLM to various evaluation criteria. Using direct assessment, we demonstrate that InfoLM achieves statistically significant improvement and over $10$ points of correlation gains in many configurations on both summarization and data2text generation.
相關內容
Medical imaging plays a pivotal role in diagnosis and treatment in clinical practice. Inspired by the significant progress in automatic image captioning, various deep learning (DL)-based methods have been proposed to generate radiology reports for medical images. Despite promising results, previous works overlook the uncertainties of their models and are thus unable to provide clinicians with the reliability/confidence of the generated radiology reports to assist their decision-making. In this paper, we propose a novel method to explicitly quantify both the visual uncertainty and the textual uncertainty for DL-based radiology report generation. Such multi-modal uncertainties can sufficiently capture the model confidence degree at both the report level and the sentence level, and thus they are further leveraged to weight the losses for more comprehensive model optimization. Experimental results have demonstrated that the proposed method for model uncertainty characterization and estimation can produce more reliable confidence scores for radiology report generation, and the modified loss function, which takes into account the uncertainties, leads to better model performance on two public radiology report datasets. In addition, the quality of the automatically generated reports was manually evaluated by human raters and the results also indicate that the proposed uncertainties can reflect the variance of clinical diagnosis.
The harmful impacts of algorithmic decision systems have recently come into focus, with many examples of systems such as machine learning (ML) models amplifying existing societal biases. Most metrics attempting to quantify disparities resulting from ML algorithms focus on differences between groups, dividing users based on demographic identities and comparing model performance or overall outcomes between these groups. However, in industry settings, such information is often not available, and inferring these characteristics carries its own risks and biases. Moreover, typical metrics that focus on a single classifier's output ignore the complex network of systems that produce outcomes in real-world settings. In this paper, we evaluate a set of metrics originating from economics, distributional inequality metrics, and their ability to measure disparities in content exposure in a production recommendation system, the Twitter algorithmic timeline. We define desirable criteria for metrics to be used in an operational setting, specifically by ML practitioners. We characterize different types of engagement with content on Twitter using these metrics, and use these results to evaluate the metrics with respect to the desired criteria. We show that we can use these metrics to identify content suggestion algorithms that contribute more strongly to skewed outcomes between users. Overall, we conclude that these metrics can be useful tools for understanding disparate outcomes in online social networks.
Training the generative models with minimal corpus is one of the critical challenges for building open-domain dialogue systems. Existing methods tend to use the meta-learning framework which pre-trains the parameters on all non-target tasks then fine-tunes on the target task. However, fine-tuning distinguishes tasks from the parameter perspective but ignores the model-structure perspective, resulting in similar dialogue models for different tasks. In this paper, we propose an algorithm that can customize a unique dialogue model for each task in the few-shot setting. In our approach, each dialogue model consists of a shared module, a gating module, and a private module. The first two modules are shared among all the tasks, while the third one will differentiate into different network structures to better capture the characteristics of the corresponding task. The extensive experiments on two datasets show that our method outperforms all the baselines in terms of task consistency, response quality, and diversity.
As a crucial component in task-oriented dialog systems, the Natural Language Generation (NLG) module converts a dialog act represented in a semantic form into a response in natural language. The success of traditional template-based or statistical models typically relies on heavily annotated data, which is infeasible for new domains. Therefore, it is pivotal for an NLG system to generalize well with limited labelled data in real applications. To this end, we present FewShotWoz, the first NLG benchmark to simulate the few-shot learning setting in task-oriented dialog systems. Further, we develop the SC-GPT model. It is pre-trained on a large set of annotated NLG corpus to acquire the controllable generation ability, and fine-tuned with only a few domain-specific labels to adapt to new domains. Experiments on FewShotWoz and the large Multi-Domain-WOZ datasets show that the proposed SC-GPT significantly outperforms existing methods, measured by various automatic metrics and human evaluations.
User engagement is a critical metric for evaluating the quality of open-domain dialogue systems. Prior work has focused on conversation-level engagement by using heuristically constructed features such as the number of turns and the total time of the conversation. In this paper, we investigate the possibility and efficacy of estimating utterance-level engagement and define a novel metric, {\em predictive engagement}, for automatic evaluation of open-domain dialogue systems. Our experiments demonstrate that (1) human annotators have high agreement on assessing utterance-level engagement scores; (2) conversation-level engagement scores can be predicted from properly aggregated utterance-level engagement scores. Furthermore, we show that the utterance-level engagement scores can be learned from data. These scores can improve automatic evaluation metrics for open-domain dialogue systems, as shown by correlation with human judgements. This suggests that predictive engagement can be used as a real-time feedback for training better dialogue models.
A robust evaluation metric has a profound impact on the development of text generation systems. A desirable metric compares system output against references based on their semantics rather than surface forms. In this paper we investigate strategies to encode system and reference texts to devise a metric that shows a high correlation with human judgment of text quality. We validate our new metric, namely MoverScore, on a number of text generation tasks including summarization, machine translation, image captioning, and data-to-text generation, where the outputs are produced by a variety of neural and non-neural systems. Our findings suggest that metrics combining contextualized representations with a distance measure perform the best. Such metrics also demonstrate strong generalization capability across tasks. For ease-of-use we make our metrics available as web service.
We propose BERTScore, an automatic evaluation metric for text generation. Analogous to common metrics, \method computes a similarity score for each token in the candidate sentence with each token in the reference. However, instead of looking for exact matches, we compute similarity using contextualized BERT embeddings. We evaluate on several machine translation and image captioning benchmarks, and show that BERTScore correlates better with human judgments than existing metrics, often significantly outperforming even task-specific supervised metrics.
Few-shot Learning aims to learn classifiers for new classes with only a few training examples per class. Existing meta-learning or metric-learning based few-shot learning approaches are limited in handling diverse domains with various number of labels. The meta-learning approaches train a meta learner to predict weights of homogeneous-structured task-specific networks, requiring a uniform number of classes across tasks. The metric-learning approaches learn one task-invariant metric for all the tasks, and they fail if the tasks diverge. We propose to deal with these limitations with meta metric learning. Our meta metric learning approach consists of task-specific learners, that exploit metric learning to handle flexible labels, and a meta learner, that discovers good parameters and gradient decent to specify the metrics in task-specific learners. Thus the proposed model is able to handle unbalanced classes as well as to generate task-specific metrics. We test our approach in the `$k$-shot $N$-way' few-shot learning setting used in previous work and new realistic few-shot setting with diverse multi-domain tasks and flexible label numbers. Experiments show that our approach attains superior performances in both settings.
Automatic summarization of natural language is a current topic in computer science research and industry, studied for decades because of its usefulness across multiple domains. For example, summarization is necessary to create reviews such as this one. Research and applications have achieved some success in extractive summarization (where key sentences are curated), however, abstractive summarization (synthesis and re-stating) is a hard problem and generally unsolved in computer science. This literature review contrasts historical progress up through current state of the art, comparing dimensions such as: extractive vs. abstractive, supervised vs. unsupervised, NLP (Natural Language Processing) vs Knowledge-based, deep learning vs algorithms, structured vs. unstructured sources, and measurement metrics such as Rouge and BLEU. Multiple dimensions are contrasted since current research uses combinations of approaches as seen in the review matrix. Throughout this summary, synthesis and critique is provided. This review concludes with insights for improved abstractive summarization measurement, with surprising implications for detecting understanding and comprehension in general.
Class labels have been empirically shown useful in improving the sample quality of generative adversarial nets (GANs). In this paper, we mathematically study the properties of the current variants of GANs that make use of class label information. With class aware gradient and cross-entropy decomposition, we reveal how class labels and associated losses influence GAN's training. Based on that, we propose Activation Maximization Generative Adversarial Networks (AM-GAN) as an advanced solution. Comprehensive experiments have been conducted to validate our analysis and evaluate the effectiveness of our solution, where AM-GAN outperforms other strong baselines and achieves state-of-the-art Inception Score (8.91) on CIFAR-10. In addition, we demonstrate that, with the Inception ImageNet classifier, Inception Score mainly tracks the diversity of the generator, and there is, however, no reliable evidence that it can reflect the true sample quality. We thus propose a new metric, called AM Score, to provide more accurate estimation on the sample quality. Our proposed model also outperforms the baseline methods in the new metric.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191