Metaphorical User Simulators for Evaluating Task-oriented Dialogue Systems

Task-oriented dialogue systems (TDSs) are assessed mainly in an offline setting or through human evaluation. The evaluation is often limited to single-turn or very time-intensive. As an alternative, user simulators that mimic user behavior allow us to consider a broad set of user goals to generate human-like conversations for simulated evaluation. Employing existing user simulators to evaluate TDSs is challenging as user simulators are primarily designed to optimize dialogue policies for TDSs and have limited evaluation capability. Moreover, the evaluation of user simulators is an open challenge. In this work, we proposes a metaphorical user simulator for endto-end TDS evaluation. We also propose a tester-based evaluation framework to generate variants, i.e., dialogue systems with different capabilities. Our user simulator constructs a metaphorical user model that assists the simulator in reasoning by referring to prior knowledge when encountering new items. We estimate the quality of simulators by checking the simulated interactions between simulators and variants. Our experiments are conducted using three TDS datasets. The metaphorical user simulator demonstrates better consistency with manual evaluation than Agenda-based simulator and Seq2seq model on three datasets; our tester framework demonstrates efficiency, and our approach demonstrates better generalization and scalability.

PRADA: Practical Black-Box Adversarial Attacks against Neural Ranking Models

Neural ranking models (NRMs) have shown remarkable success in recent years, especially with pre-trained language models. However, deep neural models are notorious for their vulnerability to adversarial examples. Adversarial attacks may become a new type of web spamming technique given our increased reliance on neural information retrieval models. Therefore, it is important to study potential adversarial attacks to identify vulnerabilities of NRMs before they are deployed. In this paper, we introduce the Adversarial Document Ranking Attack (ADRA) task against NRMs, which aims to promote a target document in rankings by adding adversarial perturbations to its text. We focus on the decision-based black-box attack setting, where the attackers have no access to the model parameters and gradients, but can only acquire the rank positions of the partial retrieved list by querying the target model. This attack setting is realistic in real-world search engines. We propose a novel Pseudo Relevance-based ADversarial ranking Attack method (PRADA) that learns a surrogate model based on Pseudo Relevance Feedback (PRF) to generate gradients for finding the adversarial perturbations. Experiments on two web search benchmark datasets show that PRADA can outperform existing attack strategies and successfully fool the NRM with small indiscernible perturbations of text.

Do Lessons from Metric Learning Generalize to Image-Caption Retrieval?

The triplet loss with semi-hard negatives has become the de facto choice for image-caption retrieval (ICR) methods that are optimized from scratch. Recent progress in metric learning has given rise to new loss functions that outperform the triplet loss on tasks such as image retrieval and representation learning. We ask whether these findings generalize to the setting of ICR by comparing three loss functions on two ICR methods. We answer this question negatively: the triplet loss with semi-hard negative mining still outperforms newly introduced loss functions from metric learning on the ICR task. To gain a better understanding of these outcomes, we introduce an analysis method to compare loss functions by counting how many samples contribute to the gradient w.r.t. the query representation during optimization. We find that loss functions that result in lower evaluation scores on the ICR task, in general, take too many (non-informative) samples into account when computing a gradient w.r.t. the query representation, which results in sub-optimal performance. The triplet loss with semi-hard negatives is shown to outperform the other loss functions, as it only takes one (hard) negative into account when computing the gradient.

Extending CLIP for Category-to-image Retrieval in E-commerce

E-commerce provides rich multimodal data that is barely leveraged in practice. One aspect of this data is a category tree that is being used in search and recommendation. However, in practice, during a user's session there is often a mismatch between a textual and a visual representation of a given category. Motivated by the problem, we introduce the task of category-to-image retrieval in e-commerce and propose a model for the task, CLIP-ITA. The model leverages information from multiple modalities (textual, visual, and attribute modality) to create product representations. We explore how adding information from multiple modalities (textual, visual, and attribute modality) impacts the model's performance. In particular, we observe that CLIP-ITA significantly outperforms a comparable model that leverages only the visual modality and a comparable model that leverages the visual and attribute modality.

Understanding and Mitigating the Effect of Outliers in Fair Ranking

Traditional ranking systems are expected to sort items in the order of their relevance and thereby maximize their utility. In fair ranking, utility is complemented with fairness as an optimization goal. Recent work on fair ranking focuses on developing algorithms to optimize for fairness, given position-based exposure. In contrast, we identify the potential of outliers in a ranking to influence exposure and thereby negatively impact fairness. An outlier in a list of items can alter the examination probabilities, which can lead to different distributions of attention, compared to position-based exposure. We formalize outlierness in a ranking, show that outliers are present in realistic datasets, and present the results of an eye-tracking study, showing that users scanning order and the exposure of items are influenced by the presence of outliers. We then introduce OMIT, a method for fair ranking in the presence of outliers. Given an outlier detection method, OMIT improves fair allocation of exposure by suppressing outliers in the top-k ranking. Using an academic search dataset, we show that outlierness optimization leads to a fairer policy that displays fewer outliers in the top-k, while maintaining a reasonable trade-off between fairness and utility.

Parameter Efficient Deep Probabilistic Forecasting

Probabilistic time series forecasting is crucial in many application domains such as retail, ecommerce, finance, or biology. With the increasing availability of large volumes of data, a number of neural architectures have been proposed for this problem. In particular, Transformer-based methods achieve state-of-the-art performance on real-world benchmarks. However, these methods require a large number of parameters to be learned, which imposes high memory requirements on the computational resources for training such models. To address this problem, we introduce a novel Bidirectional Temporal Convolutional Network (BiTCN), which requires an order of magnitude less parameters than a common Transformer-based approach. Our model combines two Temporal Convolutional Networks (TCNs): the first network encodes future covariates of the time series, whereas the second network encodes past observations and covariates. We jointly estimate the parameters of an output distribution via these two networks. Experiments on four real-world datasets show that our method performs on par with four state-of-the-art probabilistic forecasting methods, including a Transformer-based approach and WaveNet, on two point metrics (sMAPE, NRMSE) as well as on a set of range metrics (quantile loss percentiles) in the majority of cases. Secondly, we demonstrate that our method requires significantly less parameters than Transformer-based methods, which means the model can be trained faster with significantly lower memory requirements, which as a consequence reduces the infrastructure cost for deploying these models.

It Is Different When Items Are Older: Debiasing Recommendations When Selection Bias and User Preferences Are Dynamic

User interactions with recommender systems (RSs) are affected by user selection bias, e.g., users are more likely to rate popular items (popularity bias) or items that they expect to enjoy beforehand (positivity bias). Methods exist for mitigating the effects of selection bias in user ratings on the evaluation and optimization of RSs. However, these methods treat selection bias as static, despite the fact that the popularity of an item may change drastically over time and the fact that user preferences may also change over time. We focus on the age of an item and its effect on selection bias and user preferences. Our experimental analysis reveals that the rating behavior of users on the MovieLens dataset is better captured by methods that consider effects from the age of item on bias and preferences. We theoretically show that in a dynamic scenario in which both the selection bias and user preferences are dynamic, existing debiasing methods are no longer unbiased. To address this limitation, we introduce DebiAsing in the dyNamiC scEnaRio (DANCER), a novel debiasing method that extends the inverse propensity scoring debiasing method to account for dynamic selection bias and user preferences. Our experimental results indicate that DANCER improves rating prediction performance compared to debiasing methods that incorrectly assume that selection bias is static in a dynamic scenario. To the best of our knowledge, DANCER is the first debiasing method that accounts for dynamic selection bias and user preferences in RSs.

Teaching Fairness, Accountability, Confidentiality, and Transparency in Artificial Intelligence through the Lens of Reproducibility

In this work we explain the setup for a technical, graduate-level course on Fairness, Accountability, Confidentiality and Transparency in Artificial Intelligence (FACT-AI) at the University of Amsterdam, which teaches FACT-AI concepts through the lens of reproducibility. The focal point of the course is a group project based on reproducing existing FACT-AI algorithms from top AI conferences, and writing a report about their experiences. In the first iteration of the course, we created an open source repository with the code implementations from the group projects. In the second iteration, we encouraged students to submit their group projects to the Machine Learning Reproducibility Challenge, which resulted in 9 reports from our course being accepted to the challenge. We reflect on our experience teaching the course over two academic years, where one year coincided with a global pandemic, and propose guidelines for teaching FACT-AI through reproducibility in graduate-level AI programs. We hope this can be a useful resource for instructors to set up similar courses at their universities in the future.