IJCAI 2022 Papers

Artificial Immune Recognition Systems (AIRS) are powerful machine learning techniques, which aim to solve real world problems. A number of AIRS versions have produced successful prediction results. Nevertheless, these methods are unable to handle the uncertainty that could spread out at any stage of the AIRS approach. This issue is considered as a huge obstacle for having accurate and effective classification outputs. Therefore, our main objective is to handle this uncertainty using the belief function theory. We opt also in this article for an optimization over the classical AIRS approaches in order to enhance the classification performance.

Recently, attention-based models for joint intent detection and slot filling have achieved state-of-the-art performance. However, we think the conventional attention can only capture the first-order feature interaction between two tasks and is insufficient. To address this issue, we propose a unified BiLinear attention block, which leverages bilinear pooling to synchronously explore both the contextual and channel-wise bilinear attention distributions to capture the second-order interactions between the input intent and slot features. Higher-order interactions are constructed by combining many such blocks and exploiting Exponential Linear activations. Furthermore, we present a Higher-order Attention Network (HAN) to jointly model them. The experimental results show that our approach outperforms the state-of-the-art results. We also conduct experiments on the new SLURP dataset, and give a discussion on HAN’s properties, i.e., robustness and generalization.

Online peer-to-peer support platforms enable conversations between millions of people who seek and provide mental health support. If successful, web-based mental health conversations could improve access to treatment and reduce the global disease burden. Psychologists have repeatedly demonstrated that empathy, the ability to understand and feel the emotions and experiences of others, is a key component leading to positive outcomes in supportive conversations. However, recent studies have shown that highly empathic conversations are rare in online mental health platforms. In this paper, we work towards improving empathy in online mental health support conversations. We introduce a new task of empathic rewriting which aims to transform low-empathy conversational posts to higher empathy. Learning such transformations is challenging and requires a deep understanding of empathy while maintaining conversation quality through text fluency and specificity to the conversational context. Here we propose Partner, a deep reinforcement learning (RL) agent that learns to make sentence-level edits to posts in order to increase the expressed level of empathy while maintaining conversation quality. Our RL agent leverages a policy network, based on a transformer language model adapted from GPT-2, which performs the dual task of generating candidate empathic sentences and adding those sentences at appropriate positions. Through a combination of automatic and human evaluation, we demonstrate that Partner successfully generates more empathic, specific, and diverse responses and outperforms NLP methods from related tasks such as style transfer and empathic dialogue generation.

Current document-level neural machine translation (NMT) systems have achieved remarkable progress with document context. Nevertheless, discourse information that has been proven effective in many NLP tasks is ignored in most previous work. In this work, we aim at incorporating the coherence information hidden within the RST-style discourse structure into machine translation. To achieve it, we propose a document-level NMT system enhanced with the discourse-aware document context, which is named Disco2NMT. Specifically, Disco2NMT models document context based on the discourse dependency structures through a hierarchical architecture. We first convert the RST tree of an article into a dependency structure and then build the graph convolutional network (GCN) upon the segmented EDUs under the guidance of RST dependencies to capture the discourse-aware context for NMT incorporation. We conduct experiments on the document-level English-German and English-Chinese translation tasks with three domains (TED, News, and Europarl). Experimental results show that our Disco2NMT model significantly surpasses both context-agnostic and context-aware baseline systems on multiple evaluation indicators.

Deep generative models, such as Variational Autoencoders (VAEs) and Generative Adversarial Networks (GANs), have been employed widely in computational creativity research. However, such models discourage out-of-distribution generation to avoid spurious sample generation, thereby limiting their creativity. Thus, incorporating research on human creativity into generative deep learning techniques presents an opportunity to make their outputs more compelling and human-like. As we see the emergence of generative models directed toward creativity research, a need for machine learning-based surrogate metrics to characterize creative output from these models is imperative. We propose group-based subset scanning to identify, quantify, and characterize creative processes by detecting a subset of anomalous node-activations in the hidden layers of the generative models. Our experiments on the standard image benchmarks and their ``creatively generated'' variants reveal that the proposed subset scores distribution is more useful for detecting novelty in creative processes in the activation space rather than the pixel space. Further, we found that creative samples generate larger subsets of anomalies than normal or non-creative samples across datasets. The node activations highlighted during the creative decoding process are different from those responsible for the normal sample generation. Lastly, we assess if the images from the subsets selected by our method were also found creative by human evaluators, presenting a link between creativity perception in humans and node activations within deep neural nets.

Infectious diseases have been recognized as major public health concerns for decades. Close contact discovery is playing an indispensable role in preventing epidemic transmission. In this light, we study the continuous exposure search problem: Given a collection of moving objects and a collection of moving queries, we continuously discover all objects that have been directly and indirectly exposed to at least one query over a period of time. Our problem targets a variety of applications, including but not limited to disease control, epidemic pre-warning, information spreading, and co-movement mining. To answer this problem, we develop an exact group processing algorithm with optimization strategies. Further, we propose an approximate algorithm that substantially improves the efficiency without false dismissal. Extensive experiments offer insight into effectiveness and efficiency of our proposed algorithms.

As online communication grows, memes have continued to evolve and circulate as succinct multimodal forms of communication. However, computational approaches applied to meme-related tasks lack the same depth and contextual sensitivity of non-computational approaches and struggle to interpret intra-modal dynamics and referentiality. This research proposes to a ‘meme genealogy’ of key features and relationships between memes to inform a knowledge base constructed from meme-specific online sources and embed connotative meaning or contextual information in memes. The proposed methods provide a basis to train contextually sensitive computational models for analysing memes and applications in semi-automated meme annotation.

It is challenging for reinforcement learning (RL) algorithms to succeed in real-world applications. Take financial trading as an example, the market information is noisy yet imperfect and the macroeconomic regulation or other factors may shift between training and evaluation, thus it requires both generalization and high sample efficiency for resolving the task. However, directly applying typical RL algorithms can lead to poor performance in such scenarios. To derive a robust and applicable RL algorithm, in this work, we design a simple but effective method named Ensemble Proximal Policy Optimization (EPPO), which learns ensemble policies in an end-to-end manner. Notably, EPPO combines each policy and the policy ensemble organically and optimizes both simultaneously. In addition, EPPO adopts a diversity enhancement regularization over the policy space which helps to generalize to unseen states and promotes exploration. We theoretically prove that EPPO can increase exploration efficacy, and through comprehensive experimental evaluations on various tasks, we demonstrate that EPPO achieves higher efficiency and is robust for real-world applications compared with vanilla policy optimization algorithms and other ensemble methods. Code and supplemental materials are available at https://seqml.github.io/eppo.

This work systematically investigates the adversarial robustness of deep image denoisers (DIDs), i.e, how well DIDs can recover the ground truth from noisy observations degraded by adversarial perturbations. Firstly, to evaluate DIDs’ robustness, we propose a novel adversarial attack, namely Observation-based Zero-mean Attack (OBSATK), to craft adversarial zero-mean perturbations on given noisy images. We find that existing DIDs are vulnerable to the adversarial noise generated by OBSATK. Secondly, to robustify DIDs, we pro- pose an adversarial training strategy, hybrid adversarial training (HAT), that jointly trains DIDs with adversarial and non-adversarial noisy data to ensure that the reconstruction quality is high and the denoisers around non-adversarial data are locally smooth. The resultant DIDs can effectively remove various types of synthetic and adversarial noise. We also uncover that the robustness of DIDs benefits their generalization capability on unseen real-world noise. Indeed, HAT-trained DIDs can recover high-quality clean images from real-world noise even without training on real noisy data. Extensive experiments on benchmark datasets, including Set68, PolyU, and SIDD, corroborate the effectiveness of OBSATK and HAT.

Advances in multi-agent reinforcement learning (MARL) enable sequential decision making for a range of exciting multi-agent applications such as cooperative AI and autonomous driving. Explaining agent decisions is crucial for improving system transparency, increasing user satisfaction, and facilitating human-agent collaboration. However, existing works on explainable reinforcement learning mostly focus on the single-agent setting and are not suitable for addressing challenges posed by multi-agent environments. We present novel methods to generate two types of policy explanations for MARL: (i) policy summarization about the agent cooperation and task sequence, and (ii) language explanations to answer queries about agent behavior. Experimental results on three MARL domains demonstrate the scalability of our methods. A user study shows that the generated explanations significantly improve user performance and increase subjective ratings on metrics such as user satisfaction.

Point cloud segmentation plays an important role in AI applications such as autonomous driving, AR, and VR. However, previous point cloud segmentation neural networks rarely pay attention to the topological correctness of the segmentation results. In this paper, focusing on the perspective of topology awareness. First, to optimize the distribution of segmented predictions from the perspective of topology, we introduce the persistent homology theory in topology into a 3D point cloud deep learning framework. Second, we propose a topology-aware 3D point cloud segmentation module, TopoSeg. Specifically, we design a topological loss function embedded in TopoSeg module, which imposes topological constraints on the segmentation of 3D point clouds. Experiments show that our proposed TopoSeg module can be easily embedded into the point cloud segmentation network and improve the segmentation performance. In addition, based on the constructed topology loss function, we propose a topology-aware point cloud edge extraction algorithm, which is demonstrated that has strong robustness.

Residential segregation in metropolitan areas is a phenomenon that can be observed all over the world. Recently, this was investigated via game-theoretic models. There, selfish agents of two types are equipped with a monotone utility function that ensures higher utility if an agent has more same-type neighbors. The agents strategically choose their location on a given graph that serves as residential area to maximize their utility. However, sociological polls suggest that real-world agents are actually favoring mixed-type neighborhoods, and hence should be modeled via non-monotone utility functions. To address this, we study Swap Schelling Games with single-peaked utility functions. Our main finding is that tolerance, i.e., agents favoring fifty-fifty neighborhoods or being in the minority, is necessary for equilibrium existence on almost regular or bipartite graphs. Regarding the quality of equilibria, we derive (almost) tight bounds on the Price of Anarchy and the Price of Stability. In particular, we show that the latter is constant on bipartite and almost regular graphs.

We present ``Together about Dementia'', a mobile health app that aims to help persons with dementia when they get lost through the help of caregivers, relatives, and volunteering citizens. When the app detects that a person with dementia is disoriented, wandering, and getting lost, it triggers an alarm that activates a relative and possibly a volunteer in the proximity. A backend system based on a microservice and serverless architecture performs the detection of wandering and the subsequent coordination of users that are put on a mission to the rescue of the person with dementia. The backend system implements an AI technique for spatio-temporal anomaly detection based on location data recorded by the frontend system installed on the portable device of the person with dementia.

Membership inference attacks (MIAs) aim to determine whether a specific sample was used to train a predictive model. Knowing this may indeed lead to a privacy breach. Most MIAs, however, make use of the model's prediction scores - the probability of each output given some input - following the intuition that the trained model tends to behave differently on its training data. We argue that this is a fallacy for many modern deep network architectures. Consequently, MIAs will miserably fail since overconfidence leads to high false-positive rates not only on known domains but also on out-of-distribution data and implicitly acts as a defense against MIAs. Specifically, using generative adversarial networks, we are able to produce a potentially infinite number of samples falsely classified as part of the training data. In other words, the threat of MIAs is overestimated, and less information is leaked than previously assumed. Moreover, there is actually a trade-off between the overconfidence of models and their susceptibility to MIAs: the more classifiers know when they do not know, making low confidence predictions, the more they reveal the training data.

Batch-Normalization (BN) layers have become fundamental components in the evermore complex deep neural network architectures. Such models require acceleration processes for deployment on edge devices. However, BN layers add computation bottlenecks due to the sequential operation processing: thus, a key, yet often overlooked component of the acceleration process is BN layers folding. In this paper, we demonstrate that the current BN folding approaches are suboptimal in terms of how many layers can be removed. We therefore provide a necessary and sufficient condition for BN folding and a corresponding optimal algorithm. The proposed approach systematically outperforms existing baselines and allows to dramatically reduce the inference time of deep neural networks.

Recent advances in deep reinforcement learning (DRL) have largely promoted the performance of adaptive traffic signal control (ATSC). Nevertheless, regarding the implementation, most works are cumbersome in terms of storage and computation. This hinders their deployment on scenarios where resources are limited. In this work, we propose TinyLight, the first DRL-based ATSC model that is designed for devices with extremely limited resources. TinyLight first constructs a super-graph to associate a rich set of candidate features with a group of light-weighted network blocks. Then, to diminish the model's resource consumption, we ablate edges in the super-graph automatically with a novel entropy-minimized objective function. This enables TinyLight to work on a standalone microcontroller with merely 2KB RAM and 32KB ROM. We evaluate TinyLight on multiple road networks with real-world traffic demands. Experiments show that even with extremely limited resources, TinyLight still achieves competitive performance. The source code and appendix of this work can be found at https://bit.ly/38hH8t8.

In participatory budgeting the stakeholders collectively decide which projects from a set of proposed projects should be implemented. This decision underlies both time and monetary constraints. In reality it is often impossible to figure out the exact cost of each project in advance, it is only known after a project is finished. To reduce risk, one can implement projects one after the other to be able to react to higher costs of a previous project. However, this will increase execution time drastically. We generalize existing frameworks to capture this setting, study desirable properties of algorithms for this problem, and show that some desirable properties are incompatible. Then we present and analyze algorithms that trade-off desirable properties.

Several hierarchical planning systems feature a rich level of language features making them capable of expressing real-world problems. One such feature that's used by several current planning systems is causal links, which are used to track search progress. The formalism combining Hierarchical Task Network (HTN) planning with these links known from Partial Order Causal Link (POCL) planning is often referred to as hybrid planning. In this paper we study the computational complexity of such hybrid planning problems. More specifically, we provide missing membership results to existing hardness proofs and thereby provide tight complexity bounds for all known subclasses of hierarchical planning problems. We also re-visit and correct a result from the literature for plan verification showing that it remains NP-complete even in the absence of a task hierarchy.

Temporal knowledge graphs (TKGs) have been widely used in various fields that model the dynamics of facts along the timeline. In the extrapolation setting of TKG reasoning, since facts happening in the future are entirely unknowable, insight into history is the key to predicting future facts. However, it is still a great challenge for existing models as they hardly learn the characteristics of historical events adequately. From the perspective of historical development laws, comprehensively considering the sequential, repetitive, and cyclical patterns of historical facts is conducive to predicting future facts. To this end, we propose a novel representation learning model for TKG reasoning, namely TiRGN, a time-guided recurrent graph network with local-global historical patterns. Specifically, TiRGN uses a local recurrent graph encoder network to model the historical dependency of events at adjacent timestamps and uses the global history encoder network to collect repeated historical facts. After the trade-off between the two encoders, the final inference is performed by a decoder with periodicity. We use six benchmark datasets to evaluate the proposed method. The experimental results show that TiRGN outperforms the state-of-the-art TKG reasoning methods in most cases.