IJCAI 2023 论文列表

The growing population has accelerated the process of urbanisation, and is putting under stress the urban transport infrastructure. This results in increased traffic congestion, with significant health, economy, and social issues. Artificial Intelligence techniques are increasingly demonstrating their capabilities in predicting and supporting urban traffic control, by extending the abilities of traffic authorities in planning and reacting to different traffic conditions. In this context, our main research topic fits in the autonomic traffic control theme, with the aim of supporting the design of autonomous traffic control systems.

When planning the tasks of some physical entities that need to perform actions in the world (e.g., a Robot) it is necessary to take into account quite complex models for ensuring that the plan is actually executable. Indeed the state of these systems evolves according to potentially non-linear dynamics where interdependent discrete and continuous changes happen over the entire course of the task. Systems of this kind are typically compactly represented in planning using languages mixing propositional logic and mathematics. However, these languages are still poorly understood and exploited. What are the difficulties for planning in these settings? How can we build systems that can scale up over realistically sized problems? What are the domains which can benefit from these languages? This short paper shows the main two ingredients that are needed to build a heuristic search planner, outline the main impact that such techniques have on application, and provide some open challenges. These models and relative planners hold the promise to deliver explainable AI solutions that do not rely on large amounts of data.

Boolean functions and their representation through logics, circuits, machine learning classifiers, or binary decision diagrams (BDDs) play a central role in the design and analysis of computing systems. Quantifying the relative impact of variables on the truth value by means of importance values can provide useful insights to steer system design and debugging. In this paper, we introduce a uniform framework for reasoning about such values, relying on a generic notion of importance value functions (IVFs). The class of IVFs is defined by axioms motivated from several notions of importance values introduced in the literature, including Ben-Or and Linial’s influence and Chockler, Halpern, and Kupferman’s notion of responsibility and blame. We establish a connection between IVFs and game-theoretic concepts such as Shapley and Banzhaf values, both of which measure the impact of players on outcomes in cooperative games. Exploiting BDD-based symbolic methods and projected model counting, we devise and evaluate practical computation schemes for IVFs.

Reaction and retrosynthesis prediction are two fundamental tasks in computational chemistry. In recent years, these two tasks have attracted great attentions from both machine learning and drug discovery communities. Various deep learning approaches have been proposed to tackle these two problems and achieved initial success. In this survey, we conduct a comprehensive investigation on advanced deep learning-based reaction and retrosynthesis prediction models. We first summarize the design mechanism, strengths and weaknesses of the state-of-the-art approaches. Then we further discuss limitations of current solutions and open challenges in the problem itself. Last but not the least, we present some promising directions to facilitate future research. To our best knowledge, this paper is the first comprehensive and systematic survey on unified understanding of reaction and retrosynthesis prediction.

Hyperbolic neural networks can effectively capture the inherent hierarchy of graph datasets, and consequently a powerful choice of GNNs. However, they entangle multiple incongruent (gyro-)vector spaces within a layer, which makes them limited in terms of generalization and scalability. In this work, we propose the Poincaré disk model as our search space, and apply all approximations on the disk (as if the disk is a tangent space derived from the origin), thus getting rid of all inter-space transformations. Such an approach enables us to propose a hyperbolic normalization layer and to further simplify the entire hyperbolic model to a Euclidean model cascaded with our hyperbolic normalization layer. We applied our proposed nonlinear hyperbolic normalization to the current state-of-the-art homogeneous and multi-relational graph networks. We demonstrate that our model not only leverages the power of Euclidean networks such as interpretability and efficient execution of various model components, but also outperforms both Euclidean and hyperbolic counterparts on various benchmarks. Our code is made publicly available at https://github.com/oom-debugger/ijcai23.

Deep learning has achieved remarkable success in learning representations for molecules, which is crucial for various biochemical applications, ranging from property prediction to drug design. However, training Deep Neural Networks (DNNs) from scratch often requires abundant labeled molecules, which are expensive to acquire in the real world. To alleviate this issue, tremendous efforts have been devoted to Chemical Pre-trained Models (CPMs), where DNNs are pre-trained using large-scale unlabeled molecular databases and then fine-tuned over specific downstream tasks. Despite the prosperity, there lacks a systematic review of this fast-growing field. In this paper, we present the first survey that summarizes the current progress of CPMs. We first highlight the limitations of training molecular representation models from scratch to motivate CPM studies. Next, we systematically review recent advances on this topic from several key perspectives, including molecular descriptors, encoder architectures, pre-training strategies, and applications. We also highlight the challenges and promising avenues for future research, providing a useful resource for both machine learning and scientific communities.

Association rule mining is one of the well-studied and most important knowledge discovery task in data mining. In this paper, we first introduce the k-disjunctive support based itemset, a generalization of the traditional model of itemset by allowing the absence of up to k items in each transaction matching the itemset. Then, to discover more expressive rules from data, we define the concept of (k, k′)-disjunctive support based association rules by considering the antecedent and the consequent of the rule as k-disjunctive and k′-disjunctive support based itemsets, respectively. Second, we provide a polynomial-time reduction of both the problems of mining k-disjunctive support based itemsets and (k, k′)-disjunctive support based association rules to the propositional satisfiability model enumeration task. Finally, we show through an extensive campaign of experiments on several popular real-life datasets the efficiency of our proposed approach

User Behavior Modeling (UBM) plays a critical role in user interest learning, which has been extensively used in recommender systems. Crucial interactive patterns between users and items have been exploited, which brings compelling improvements in many recommendation tasks. In this paper, we attempt to provide a thorough survey of this research topic. We start by reviewing the research background of UBM. Then, we provide a systematic taxonomy of existing UBM research works, which can be categorized into four different directions including Conventional UBM, Long-Sequence UBM, Multi-Type UBM, and UBM with Side Information. Within each direction, representative models and their strengths and weaknesses are comprehensively discussed. Besides, we elaborate on the industrial practices of UBM methods with the hope of providing insights into the application value of existing UBM solutions. Finally, we summarize the survey and discuss the future prospects of this field.

Proactive dialogue systems, related to a wide range of real-world conversational applications, equip the conversational agent with the capability of leading the conversation direction towards achieving pre-defined targets or fulfilling certain goals from the system side. It is empowered by advanced techniques to progress to more complicated tasks that require strategical and motivational interactions. In this survey, we provide a comprehensive overview of the prominent problems and advanced designs for conversational agent's proactivity in different types of dialogues. Furthermore, we discuss challenges that meet the real-world application needs but require a greater research focus in the future. We hope that this first survey of proactive dialogue systems can provide the community with a quick access and an overall picture to this practical problem, and stimulate more progresses on conversational AI to the next level.

Adversarial robustness, domain generalization and dataset biases are three active lines of research contributing to out-of-distribution (OOD) evaluation on neural NLP models. However, a comprehensive, integrated discussion of the three research lines is still lacking in the literature. This survey will 1) compare the three lines of research under a unifying definition; 2) summarize their data-generating processes and evaluation protocols for each line of research; and 3) emphasize the challenges and opportunities for future work.

With the increasing popularity of masked autoencoders, self-supervised learning (SSL) in vision undertakes a similar trajectory as in NLP. Specifically, generative pretext tasks with the masked prediction have become a de facto standard SSL practice in NLP (e.g., BERT). By contrast, early attempts at generative methods in vision have been outperformed by their discriminative counterparts (like contrastive learning). However, the success of masked image modeling has revived the autoencoder-based visual pretraining method. As a milestone to bridge the gap with BERT in NLP, masked autoencoder in vision has attracted unprecedented attention. This work conducts a survey on masked autoencoders for visual SSL.

The widespread adoption of Machine Learning systems, especially in more decision-critical applications such as criminal sentencing and bank loans, has led to increased concerns about fairness implications. Algorithms and metrics have been developed to mitigate and measure these discriminations. More recently, works have identified a more challenging form of bias called intersectional bias, which encompasses multiple sensitive attributes, such as race and gender, together. In this survey, we review the state-of-the-art in intersectional fairness. We present a taxonomy for intersectional notions of fairness and mitigation. Finally, we identify the key challenges and provide researchers with guidelines for future directions.

Recent advances in Transformers have come with a huge requirement on computing resources, highlighting the importance of developing efficient training techniques to make Transformer training faster, at lower cost, and to higher accuracy by the efficient use of computation and memory resources. This survey provides the first systematic overview of the efficient training of Transformers, covering the recent progress in acceleration arithmetic and hardware, with a focus on the former. We analyze and compare methods that save computation and memory costs for intermediate tensors during training, together with techniques on hardware/algorithm co-design. We finally discuss challenges and promising areas for future research.

Dataset distillation is attracting more attention in machine learning as training sets continue to grow and the cost of training state-of-the-art models becomes increasingly high. By synthesizing datasets with high information density, dataset distillation offers a range of potential applications, including support for continual learning, neural architecture search, and privacy protection. Despite recent advances, we lack a holistic understanding of the approaches and applications. Our survey aims to bridge this gap by first proposing a taxonomy of dataset distillation, characterizing existing approaches, and then systematically reviewing the data modalities, and related applications. In addition, we summarize the challenges and discuss future directions for this field of research.

Algorithm configuration (AC) is concerned with the automated search of the most suitable parameter configuration of a parametrized algorithm. There are currently a wide variety of AC problem variants and methods proposed in the literature. Existing reviews do not take into account all derivatives of the AC problem, nor do they offer a complete classification scheme. To this end, we introduce taxonomies to describe the AC problem and features of configuration methods, respectively. Existing AC literature is classified and characterized by the provided taxonomies.

In traditional machine learning, it is trivial to conduct model evaluation since all data samples are managed centrally by a server. However, model evaluation becomes a challenging problem in federated learning (FL), which is called federated evaluation in this work. This is because clients do not expose their original data to preserve data privacy. Federated evaluation plays a vital role in client selection, incentive mechanism design, malicious attack detection, etc. In this paper, we provide the first comprehensive survey of existing federated evaluation methods. Moreover, we explore various applications of federated evaluation for enhancing FL performance and finally present future research directions by envisioning some challenges.

Recent studies on pedestrian attribute recognition progress with either explicit or implicit modeling of the co-occurence among attributes. Considering that this known a prior is highly variable and unforeseeable regarding the specific scenarios, we show that current methods can actually suffer in generalizing such fitted attributes interdependencies onto scenes or identities off the dataset distribution, resulting in the underlined bias of attributes co-occurence. To render models robust in realistic scenes, we propose the attributes-disentangled feature learning to ensure the recognition of an attribute not inferring on the existence of others, and which is sequentially formulated as a problem of mutual information minimization. Rooting from it, practical strategies are devised to efficiently decouple attributes, which substantially improve the baseline and establish state-of-the-art performance on realistic datasets like PETAzs and RAPzs.

We present a novel rule-based semantics for causal reasoning as well as a number of modal languages interpreted over it. They enable us to represent some fundamental concepts in the theory of causality including causal necessity and possibility, interventionist conditionals and Lewisian conditionals. We provide complexity results for the satisfiability checking and model checking problem for these modal languages. Moreover, we study the relationship between our rule-based semantics and the structural equation modeling (SEM) approach to causal reasoning, as well as between our rule-based semantics for causal conditionals and the standard semantics for belief base change.

We consider the problem of risk-aware Markov Decision Processes (MDPs) for Safe AI. We introduce a theoretical framework, Extended Markov Ratio Decision Processes (EMRDP), that incorporates risk into MDPs and embeds environment learning into this framework. We propose an algorithm to find the optimal policy for EMRDP with theoretical guarantees. Under a certain monotonicity assumption, this algorithm runs in strongly-polynomial time both in the discounted and expected average reward models. We validate our algorithm empirically on a Grid World benchmark, evaluating its solution quality, required number of steps, and numerical stability. We find its solution quality to be stable under data noising, while its required number of steps grows with added noise. We observe its numerical stability compared to global methods.