Exemplar-based generative models for open-domain conversation produce responses based on the exemplars provided by the retriever, taking advantage of generative models and retrieval models. However, they often ignore the retrieved exemplars while generating responses or produce responses over-fitted to the retrieved exemplars. In this paper, we argue that these drawbacks are derived from the one-to-many problem of the open-domain conversation. When the retrieved exemplar is relevant to the given context yet significantly different from the gold response, the exemplar-based generative models are trained to ignore the exemplar since the exemplar is not helpful for generating the gold response. On the other hand, when the retrieved exemplar is lexically similar to the gold response, the generative models are trained to rely on the exemplar highly. Therefore, we propose a training method selecting exemplars that are semantically relevant to the gold response but lexically distanced from the gold response to mitigate the above disadvantages. In the training phase, our proposed training method first uses the gold response instead of dialogue context as a query to select exemplars that are semantically relevant to the gold response. And then, it eliminates the exemplars that lexically resemble the gold responses to alleviate the dependency of the generative models on that exemplars. The remaining exemplars could be irrelevant to the given context since they are searched depending on the gold response. Thus, our proposed training method further utilizes the relevance scores between the given context and the exemplars to penalize the irrelevant exemplars. Extensive experiments demonstrate that our proposed training method alleviates the drawbacks of the existing exemplar-based generative models and significantly improves the performance in terms of appropriateness and informativeness.
Scene graph generation (SGG) aims to capture a wide variety of interactions between pairs of objects, which is essential for full scene understanding. Existing SGG methods trained on the entire set of relations fail to acquire complex reasoning about visual and textual correlations due to various biases in training data. Learning on trivial relations that indicate generic spatial configuration like 'on' instead of informative relations such as 'parked on' does not enforce this complex reasoning, harming generalization. To address this problem, we propose a novel framework for SGG training that exploits relation labels based on their informativeness. Our model-agnostic training procedure imputes missing informative relations for less informative samples in the training data and trains a SGG model on the imputed labels along with existing annotations. We show that this approach can successfully be used in conjunction with state-of-the-art SGG methods and improves their performance significantly in multiple metrics on the standard Visual Genome benchmark. Furthermore, we obtain considerable improvements for unseen triplets in a more challenging zero-shot setting.
Small and medium enterprises rely on detailed Web analytics to be informed about their market and competition. Focused crawlers meet this demand by crawling and indexing specific parts of the Web. Critically, a focused crawler must quickly find new pages that have not yet been indexed. Since a new page can be discovered only by following a new outlink, predicting new outlinks is very relevant in practice. In the literature, many feature designs have been proposed for predicting changes in the Web. In this work we provide a structured analysis of this problem, using new outlinks as our running prediction target. Specifically, we unify earlier feature designs in a taxonomic arrangement of features along two dimensions: static versus dynamic features, and features of a page versus features of the network around it. Within this taxonomy, complemented by our new (mainly, dynamic network) features, we identify best predictors for new outlinks. Our main conclusion is that most informative features are the recent history of new outlinks on a page itself, and of its content-related pages. Hence, we propose a new 'look back, look around' (LBLA) model, that uses only these features. With the obtained predictions, we design a number of scoring functions to guide a focused crawler to pages with most new outlinks, and compare their performance. The LBLA approach proved extremely effective, outperforming other models including those that use a most complete set of features. One of the learners we use, is the recent NGBoost method that assumes a Poisson distribution for the number of new outlinks on a page, and learns its parameters. This connects the two so far unrelated avenues in the literature: predictions based on features of a page, and those based on probabilistic modelling. All experiments were carried out on an original dataset, made available by a commercial focused crawler.
Learned representations in deep reinforcement learning (DRL) have to extract task-relevant information from complex observations, balancing between robustness to distraction and informativeness to the policy. Such stable and rich representations, often learned via modern function approximation techniques, can enable practical application of the policy improvement theorem, even in high-dimensional continuous state-action spaces. Bisimulation metrics offer one solution to this representation learning problem, by collapsing functionally similar states together in representation space, which promotes invariance to noise and distractors. In this work, we generalize value function approximation bounds for on-policy bisimulation metrics to non-optimal policies and approximate environment dynamics. Our theoretical results help us identify embedding pathologies that may occur in practical use. In particular, we find that these issues stem from an underconstrained dynamics model and an unstable dependence of the embedding norm on the reward signal in environments with sparse rewards. Further, we propose a set of practical remedies: (i) a norm constraint on the representation space, and (ii) an extension of prior approaches with intrinsic rewards and latent space regularization. Finally, we provide evidence that the resulting method is not only more robust to sparse reward functions, but also able to solve challenging continuous control tasks with observational distractions, where prior methods fail.
I establish a translation invariance property of the Blackwell order over experiments, show that garbling experiments bring them closer together, and use these facts to define a cardinal measure of informativeness. Experiment $A$ is inf-norm more informative (INMI) than experiment $B$ if the infinity norm of the difference between a perfectly informative structure and $A$ is less than the corresponding difference for $B$. The better experiment is "closer" to the fully revealing experiment; distance from the identity matrix is interpreted as a measure of informativeness. This measure coincides with Blackwell's order whenever possible, is complete, order invariant, and prior-independent, making it an attractive and computationally simple extension of the Blackwell order to economic contexts.
The task of image captioning aims to generate captions directly from images via the automatically learned cross-modal generator. To build a well-performing generator, existing approaches usually need a large number of described images, which requires a huge effects on manual labeling. However, in real-world applications, a more general scenario is that we only have limited amount of described images and a large number of undescribed images. Therefore, a resulting challenge is how to effectively combine the undescribed images into the learning of cross-modal generator. To solve this problem, we propose a novel image captioning method by exploiting the Cross-modal Prediction and Relation Consistency (CPRC), which aims to utilize the raw image input to constrain the generated sentence in the commonly semantic space. In detail, considering that the heterogeneous gap between modalities always leads to the supervision difficulty of using the global embedding directly, CPRC turns to transform both the raw image and corresponding generated sentence into the shared semantic space, and measure the generated sentence from two aspects: 1) Prediction consistency. CPRC utilizes the prediction of raw image as soft label to distill useful supervision for the generated sentence, rather than employing the traditional pseudo labeling; 2) Relation consistency. CPRC develops a novel relation consistency between augmented images and corresponding generated sentences to retain the important relational knowledge. In result, CPRC supervises the generated sentence from both the informativeness and representativeness perspectives, and can reasonably use the undescribed images to learn a more effective generator under the semi-supervised scenario.
Natural language processing (NLP) systems are increasingly trained to generate open-ended text rather than classifying between responses. This makes research on evaluation metrics for generated language -- functions that score system output given the context and/or human reference responses -- of critical importance. However, different metrics have different strengths and biases, and reflect human intuitions better on some tasks than others. There is currently no simple, unified way to compare, analyse or evaluate metrics across a representative set of tasks. Here, we describe the Benchmark to Evaluate Automatic Metrics (BEAMetrics), a resource to make research into new metrics itself easier to evaluate. BEAMetrics users can quickly compare existing and new metrics with human judgements across a diverse set of tasks, quality dimensions (fluency vs. coherence vs. informativeness etc), and languages. As generation experts might predict, BEAMetrics reveals stark task-dependent differences between existing metrics, and consistently poor performance on tasks with complex answer spaces or high reliance on general knowledge. While this analysis highlights a critical issue facing current research practice, BEAMetrics also contribute to its resolution by facilitating research into better metrics -- particularly those that can account for the complex interaction between context and general knowledge inherent to many modern NLP applications. BEAMetrics is available under the MIT License: https://github.com/ThomasScialom/BEAMetrics
Hallucination is a known issue for neural abstractive summarization models. Recent work suggests that the degree of hallucination may depend on errors in the training data. In this work, we propose a new method called Contrastive Parameter Ensembling (CaPE) to use training data more effectively, utilizing variations in noise in training samples to reduce hallucination. We first select clean and noisy subsets from the training data using different automatic factual metrics. Then, we fine-tune a base summarization model, which is trained on all training samples, on the clean (noisy) subset to obtain an \textit{expert} (\textit{anti-expert}) model. Finally, we adjust the parameters of base model by the difference between parameters of the \textit{expert} and \textit{anti-expert} models, steering the base model towards the \textit{expert} model and away from the \textit{anti-expert} model. Experimental results show that CaPE improves performance across different automatic factual metrics and human evaluation, with the maximum improvement of 16.69\% and 15.78\% on summary-level dependency-arc entailment accuracy for the XSUM and CNN/DM datasets. The improvement in factual performance does not degrade the performance on other metrics of informativeness such as ROUGE.
Most online markets establish reputation systems to assist building trust between sellers and buyers. Sellers' reputations not only provide guidelines for buyers but may also inform sellers their optimal pricing strategy. In this research, we assumed two types of buyer: informed buyers and uninformed buyers. Informed buyers know more about the reputation about the seller but may incur a search cost. Then we developed a benchmark model and a competition model. We found that high reputation sellers and low reputation sellers adapt different pricing strategy depending on the informativeness of buyers and the competition among sellers. With a large proportion of informed buyers, high reputation sellers may charge lower price than low reputation sellers, which exists a negative price premium effect, in contrast to conclusions of some previous studies. Empirical findings were in consistence with our theoretical models. We collected data of five categories of products, televisions, laptops, cosmetics, shoes, and beverages, from Taobao, a leading C2C Chinese online market. Negative price premium effect was observed for TVs, laptops, and cosmetics; price premium effect was observed for beverages; no significant trend was observed for shoes. We infer product value and market complexity are the main factors of buyer informativeness.
In cooperative multi-agent reinforcement learning (MARL), where agents only have access to partial observations, efficiently leveraging local information is critical. During long-time observations, agents can build \textit{awareness} for teammates to alleviate the problem of partial observability. However, previous MARL methods usually neglect this kind of utilization of local information. To address this problem, we propose a novel framework, multi-agent \textit{Local INformation Decomposition for Awareness of teammates} (LINDA), with which agents learn to decompose local information and build awareness for each teammate. We model the awareness as stochastic random variables and perform representation learning to ensure the informativeness of awareness representations by maximizing the mutual information between awareness and the actual trajectory of the corresponding agent. LINDA is agnostic to specific algorithms and can be flexibly integrated to different MARL methods. Sufficient experiments show that the proposed framework learns informative awareness from local partial observations for better collaboration and significantly improves the learning performance, especially on challenging tasks.
While learning from synthetic training data has recently gained an increased attention, in real-world robotic applications, there are still performance deficiencies due to the so-called Sim-to-Real gap. In practice, this gap is hard to resolve with only synthetic data. Therefore, we focus on an efficient acquisition of real data within a Sim-to-Real learning pipeline. Concretely, we employ deep Bayesian active learning to minimize manual annotation efforts and devise an autonomous learning paradigm to select the data that is considered useful for the human expert to annotate. To achieve this, a Bayesian Neural Network (BNN) object detector providing reliable uncertainty estimates is adapted to infer the informativeness of the unlabeled data. Furthermore, to cope with mis-alignments of the label distribution in uncertainty-based sampling, we develop an effective randomized sampling strategy that performs favorably compared to other complex alternatives. In our experiments on object classification and detection, we show benefits of our approach and provide evidence that labeling efforts can be reduced significantly. Finally, we demonstrate the practical effectiveness of this idea in a grasping task on an assistive robot.
Automatic text summarization aims to cut down readers time and cognitive effort by reducing the content of a text document without compromising on its essence. Ergo, informativeness is the prime attribute of document summary generated by an algorithm, and selecting sentences that capture the essence of a document is the primary goal of extractive document summarization. In this paper, we employ Shannon entropy to capture informativeness of sentences. We employ Non-negative Matrix Factorization (NMF) to reveal probability distributions for computing entropy of terms, topics, and sentences in latent space. We present an information theoretic interpretation of the computed entropy, which is the bedrock of the proposed E-Summ algorithm, an unsupervised method for extractive document summarization. The algorithm systematically applies information theoretic principle for selecting informative sentences from important topics in the document. The proposed algorithm is generic and fast, and hence amenable to use for summarization of documents in real time. Furthermore, it is domain-, collection-independent and agnostic to the language of the document. Benefiting from strictly positive NMF factor matrices, E-Summ algorithm is transparent and explainable too. We use standard ROUGE toolkit for performance evaluation of the proposed method on four well known public data-sets. We also perform quantitative assessment of E-Summ summary quality by computing its semantic similarity w.r.t the original document. Our investigation reveals that though using NMF and information theoretic approach for document summarization promises efficient, explainable, and language independent text summarization, it needs to be bolstered to match the performance of deep neural methods.
When people receive new information, sometimes they revise their beliefs too much, and sometimes too little. In this paper, we show that a key driver of whether people overinfer or underinfer is the strength of the information. Based on a model in which people know which direction to update in, but not exactly how much to update, we hypothesize that people will overinfer from weak signals and underinfer from strong signals. We then test this hypothesis across four different environments: abstract experiments, a naturalistic experiment, sports betting markets, and financial markets. In each environment, our consistent and robust finding is overinference from weak signals and underinference from strong signals. Our framework and findings can help harmonize apparently contradictory results from the experimental and empirical literatures.
This paper presents an automatic method to evaluate the naturalness of natural language generation in dialogue systems. While this task was previously rendered through expensive and time-consuming human labor, we present this novel task of automatic naturalness evaluation of generated language. By fine-tuning the BERT model, our proposed naturalness evaluation method shows robust results and outperforms the baselines: support vector machines, bi-directional LSTMs, and BLEURT. In addition, the training speed and evaluation performance of naturalness model are improved by transfer learning from quality and informativeness linguistic knowledge.
Recently, Deep Neural Networks (DNNs) have made remarkable progress for text classification, which, however, still require a large number of labeled data. To train high-performing models with the minimal annotation cost, active learning is proposed to select and label the most informative samples, yet it is still challenging to measure informativeness of samples used in DNNs. In this paper, inspired by piece-wise linear interpretability of DNNs, we propose a novel Active Learning with DivErse iNterpretations (ALDEN) approach. With local interpretations in DNNs, ALDEN identifies linearly separable regions of samples. Then, it selects samples according to their diversity of local interpretations and queries their labels. To tackle the text classification problem, we choose the word with the most diverse interpretations to represent the whole sentence. Extensive experiments demonstrate that ALDEN consistently outperforms several state-of-the-art deep active learning methods.
Recent work has demonstrated the viability of using crowdsourcing as a tool for evaluating the truthfulness of public statements. Under certain conditions such as: (1) having a balanced set of workers with different backgrounds and cognitive abilities; (2) using an adequate set of mechanisms to control the quality of the collected data; and (3) using a coarse grained assessment scale, the crowd can provide reliable identification of fake news. However, fake news are a subtle matter: statements can be just biased ("cherrypicked"), imprecise, wrong, etc. and the unidimensional truth scale used in existing work cannot account for such differences. In this paper we propose a multidimensional notion of truthfulness and we ask the crowd workers to assess seven different dimensions of truthfulness selected based on existing literature: Correctness, Neutrality, Comprehensibility, Precision, Completeness, Speaker's Trustworthiness, and Informativeness. We deploy a set of quality control mechanisms to ensure that the thousands of assessments collected on 180 publicly available fact-checked statements distributed over two datasets are of adequate quality, including a custom search engine used by the crowd workers to find web pages supporting their truthfulness assessments. A comprehensive analysis of crowdsourced judgments shows that: (1) the crowdsourced assessments are reliable when compared to an expert-provided gold standard; (2) the proposed dimensions of truthfulness capture independent pieces of information; (3) the crowdsourcing task can be easily learned by the workers; and (4) the resulting assessments provide a useful basis for a more complete estimation of statement truthfulness.
Active learning aims to select samples to be annotated that yield the largest performance improvement for the learning algorithm. Many methods approach this problem by measuring the informativeness of samples and do this based on the certainty of the network predictions for samples. However, it is well-known that neural networks are overly confident about their prediction and are therefore an untrustworthy source to assess sample informativeness. In this paper, we propose a new informativeness-based active learning method. Our measure is derived from the learning dynamics of a neural network. More precisely we track the label assignment of the unlabeled data pool during the training of the algorithm. We capture the learning dynamics with a metric called label-dispersion, which is low when the network consistently assigns the same label to the sample during the training of the network and high when the assigned label changes frequently. We show that label-dispersion is a promising predictor of the uncertainty of the network, and show on two benchmark datasets that an active learning algorithm based on label-dispersion obtains excellent results.
Knowledge-grounded dialogue systems are intended to convey information that is based on evidence provided in a given source text. We discuss the challenges of training a generative neural dialogue model for such systems that is controlled to stay faithful to the evidence. Existing datasets contain a mix of conversational responses that are faithful to selected evidence as well as more subjective or chit-chat style responses. We propose different evaluation measures to disentangle these different styles of responses by quantifying the informativeness and objectivity. At training time, additional inputs based on these evaluation measures are given to the dialogue model. At generation time, these additional inputs act as stylistic controls that encourage the model to generate responses that are faithful to the provided evidence. We also investigate the usage of additional controls at decoding time using resampling techniques. In addition to automatic metrics, we perform a human evaluation study where raters judge the output of these controlled generation models to be generally more objective and faithful to the evidence compared to baseline dialogue systems.
In this paper, we answer the question of when inserting label noise (less informative labels) can instead return us more accurate and fair models. We are primarily inspired by three observations: 1) In contrast to reducing label noise rates, increasing the noise rates is easy to implement; 2) Increasing a certain class of instances' label noise to balance the noise rates (increasing-to-balancing) results in an easier learning problem; 3) Increasing-to-balancing improves fairness guarantees against label bias. In this paper, we first quantify the trade-offs introduced by increasing a certain group of instances' label noise rate w.r.t. the loss of label informativeness and the lowered learning difficulties. We analytically demonstrate when such an increase is beneficial, in terms of either improved generalization power or the fairness guarantees. Then we present a method to insert label noise properly for the task of learning with noisy labels, either without or with a fairness constraint. The primary technical challenge we face is due to the fact that we would not know which data instances are suffering from higher noise, and we would not have the ground truth labels to verify any possible hypothesis. We propose a detection method that informs us which group of labels might suffer from higher noise without using ground truth labels. We formally establish the effectiveness of the proposed solution and demonstrate it with extensive experiments.
Active learning aims to achieve greater accuracy with less training data by selecting the most useful data samples from which it learns. Single-criterion based methods (i.e., informativeness and representativeness based methods) are simple and efficient; however, they lack adaptability to different real-world scenarios. In this paper, we introduce a multiple-criteria based active learning algorithm, which incorporates three complementary criteria, i.e., informativeness, representativeness and diversity, to make appropriate selections in the active learning rounds under different data types. We consider the selection process as a Determinantal Point Process, which good balance among these criteria. We refine the query selection strategy by both selecting the hardest unlabeled data sample and biasing towards the classifiers that are more suitable for the current data distribution. In addition, we also consider the dependencies and relationships between these data points in data selection by means of centroidbased clustering approaches. Through evaluations on synthetic and real-world datasets, we show that our method performs significantly better and is more stable than other multiple-criteria based AL algorithms.