Self-Supervised Bot Play for Conversational Recommendation with Justifications

Conversational recommender systems offer the promise of interactive, engaging ways for users to find items they enjoy. We seek to improve conversational recommendation via three dimensions: 1) We aim to mimic a common mode of human interaction for recommendation: experts justify their suggestions, a seeker explains why they don't like the item, and both parties iterate through the dialog to find a suitable item. 2) We leverage ideas from conversational critiquing to allow users to flexibly interact with natural language justifications by critiquing subjective aspects. 3) We adapt conversational recommendation to a wider range of domains where crowd-sourced ground truth dialogs are not available. We develop a new two-part framework for training conversational recommender systems. First, we train a recommender system to jointly suggest items and justify its reasoning with subjective aspects. We then fine-tune this model to incorporate iterative user feedback via self-supervised bot-play. Experiments on three real-world datasets demonstrate that our system can be applied to different recommendation models across diverse domains to achieve superior performance in conversational recommendation compared to state-of-the-art methods. We also evaluate our model on human users, showing that systems trained under our framework provide more useful, helpful, and knowledgeable recommendations in warm- and cold-start settings.

Rethink, Revisit, Revise: A Spiral Reinforced Self-Revised Network for Zero-Shot Learning

Current approaches to Zero-Shot Learning (ZSL) struggle to learn generalizable semantic knowledge capable of capturing complex correlations. Inspired by \emph{Spiral Curriculum}, which enhances learning processes by revisiting knowledge, we propose a form of spiral learning which revisits visual representations based on a sequence of attribute groups (e.g., a combined group of \emph{color} and \emph{shape}). Spiral learning aims to learn generalized local correlations, enabling models to gradually enhance global learning and thus understand complex correlations. Our implementation is based on a 2-stage \emph{Reinforced Self-Revised (RSR)} framework: \emph{preview} and \emph{review}. RSR first previews visual information to construct diverse attribute groups in a weakly-supervised manner. Then, it spirally learns refined localities based on attribute groups and uses localities to revise global semantic correlations. Our framework outperforms state-of-the-art algorithms on four benchmark datasets in both zero-shot and generalized zero-shot settings, which demonstrates the effectiveness of spiral learning in learning generalizable and complex correlations. We also conduct extensive analysis to show that attribute groups and reinforced decision processes can capture complementary semantic information to improve predictions and aid explainability.

An Entropy-guided Reinforced Partial Convolutional Network for Zero-Shot Learning

Zero-Shot Learning (ZSL) aims to transfer learned knowledge from observed classes to unseen classes via semantic correlations. A promising strategy is to learn a global-local representation that incorporates global information with extra localities (i.e., small parts/regions of inputs). However, existing methods discover localities based on explicit features without digging into the inherent properties and relationships among regions. In this work, we propose a novel Entropy-guided Reinforced Partial Convolutional Network (ERPCNet), which extracts and aggregates localities progressively based on semantic relevance and visual correlations without human-annotated regions. ERPCNet uses reinforced partial convolution and entropy guidance; it not only discovers global-cooperative localities dynamically but also converges faster for policy gradient optimization. We conduct extensive experiments to demonstrate ERPCNet's performance through comparisons with state-of-the-art methods under ZSL and Generalized Zero-Shot Learning (GZSL) settings on four benchmark datasets. We also show ERPCNet is time efficient and explainable through visualization analysis.

Locality-Sensitive Experience Replay for Online Recommendation

Online recommendation requires handling rapidly changing user preferences. Deep reinforcement learning (DRL) is gaining interest as an effective means of capturing users' dynamic interest during interactions with recommender systems. However, it is challenging to train a DRL agent, due to large state space (e.g., user-item rating matrix and user profiles), action space (e.g., candidate items), and sparse rewards. Existing studies encourage the agent to learn from past experience via experience replay (ER). They adapt poorly to the complex environment of online recommender systems and are inefficient in determining an optimal strategy from past experience. To address these issues, we design a novel state-aware experience replay model, which uses locality-sensitive hashing to map high dimensional data into low-dimensional representations and a prioritized reward-driven strategy to replay more valuable experience at a higher chance. Our model can selectively pick the most relevant and salient experiences and recommend the agent with the optimal policy. Experiments on three online simulation platforms demonstrate our model' feasibility and superiority toseveral existing experience replay methods.

Weakly Supervised Contrastive Learning for Chest X-Ray Report Generation

Radiology report generation aims at generating descriptive text from radiology images automatically, which may present an opportunity to improve radiology reporting and interpretation. A typical setting consists of training encoder-decoder models on image-report pairs with a cross entropy loss, which struggles to generate informative sentences for clinical diagnoses since normal findings dominate the datasets. To tackle this challenge and encourage more clinically-accurate text outputs, we propose a novel weakly supervised contrastive loss for medical report generation. Experimental results demonstrate that our method benefits from contrasting target reports with incorrect but semantically-close ones. It outperforms previous work on both clinical correctness and text generation metrics for two public benchmarks.

Detect and Perturb: Neutral Rewriting of Biased and Sensitive Text via Gradient-based Decoding

Written language carries explicit and implicit biases that can distract from meaningful signals. For example, letters of reference may describe male and female candidates differently, or their writing style may indirectly reveal demographic characteristics. At best, such biases distract from the meaningful content of the text; at worst they can lead to unfair outcomes. We investigate the challenge of re-generating input sentences to 'neutralize' sensitive attributes while maintaining the semantic meaning of the original text (e.g. is the candidate qualified?). We propose a gradient-based rewriting framework, Detect and Perturb to Neutralize (DEPEN), that first detects sensitive components and masks them for regeneration, then perturbs the generation model at decoding time under a neutralizing constraint that pushes the (predicted) distribution of sensitive attributes towards a uniform distribution. Our experiments in two different scenarios show that DEPEN can regenerate fluent alternatives that are neutral in the sensitive attribute while maintaining the semantics of other attributes.

Modeling Dynamic Attributes for Next Basket Recommendation

Traditional approaches to next item and next basket recommendation typically extract users' interests based on their past interactions and associated static contextual information (e.g. a user id or item category). However, extracted interests can be inaccurate and become obsolete. Dynamic attributes, such as user income changes, item price changes (etc.), change over time. Such dynamics can intrinsically reflect the evolution of users' interests. We argue that modeling such dynamic attributes can boost recommendation performance. However, properly integrating them into user interest models is challenging since attribute dynamics can be diverse such as time-interval aware, periodic patterns (etc.), and they represent users' behaviors from different perspectives, which can happen asynchronously with interactions. Besides dynamic attributes, items in each basket contain complex interdependencies which might be beneficial but nontrivial to effectively capture. To address these challenges, we propose a novel Attentive network to model Dynamic attributes (named AnDa). AnDa separately encodes dynamic attributes and basket item sequences. We design a periodic aware encoder to allow the model to capture various temporal patterns from dynamic attributes. To effectively learn useful item relationships, intra-basket attention module is proposed. Experimental results on three real-world datasets demonstrate that our method consistently outperforms the state-of-the-art.

A Survey of Deep Reinforcement Learning in Recommender Systems: A Systematic Review and Future Directions

In light of the emergence of deep reinforcement learning (DRL) in recommender systems research and several fruitful results in recent years, this survey aims to provide a timely and comprehensive overview of the recent trends of deep reinforcement learning in recommender systems. We start with the motivation of applying DRL in recommender systems. Then, we provide a taxonomy of current DRL-based recommender systems and a summary of existing methods. We discuss emerging topics and open issues, and provide our perspective on advancing the domain. This survey serves as introductory material for readers from academia and industry into the topic and identifies notable opportunities for further research.

Beyond Preserved Accuracy: Evaluating Loyalty and Robustness of BERT Compression

Recent studies on compression of pretrained language models (e.g., BERT) usually use preserved accuracy as the metric for evaluation. In this paper, we propose two new metrics, label loyalty and probability loyalty that measure how closely a compressed model (i.e., student) mimics the original model (i.e., teacher). We also explore the effect of compression with regard to robustness under adversarial attacks. We benchmark quantization, pruning, knowledge distillation and progressive module replacing with loyalty and robustness. By combining multiple compression techniques, we provide a practical strategy to achieve better accuracy, loyalty and robustness.

Black-Box Attacks on Sequential Recommenders via Data-Free Model Extraction

We investigate whether model extraction can be used to "steal" the weights of sequential recommender systems, and the potential threats posed to victims of such attacks. This type of risk has attracted attention in image and text classification, but to our knowledge not in recommender systems. We argue that sequential recommender systems are subject to unique vulnerabilities due to the specific autoregressive regimes used to train them. Unlike many existing recommender attackers, which assume the dataset used to train the victim model is exposed to attackers, we consider a data-free setting, where training data are not accessible. Under this setting, we propose an API-based model extraction method via limited-budget synthetic data generation and knowledge distillation. We investigate state-of-the-art models for sequential recommendation and show their vulnerability under model extraction and downstream attacks. We perform attacks in two stages. (1) Model extraction: given different types of synthetic data and their labels retrieved from a black-box recommender, we extract the black-box model to a white-box model via distillation. (2) Downstream attacks: we attack the black-box model with adversarial samples generated by the white-box recommender. Experiments show the effectiveness of our data-free model extraction and downstream attacks on sequential recommenders in both profile pollution and data poisoning settings.