IJCAI 2020 Papers

Predicting the price movement of finance securities like stocks is an important but challenging task, due to the uncertainty of financial markets. In this paper, we propose a novel approach based on the Transformer to tackle the stock movement prediction task. Furthermore, we present several enhancements for the proposed basic Transformer. Firstly, we propose a Multi-Scale Gaussian Prior to enhance the locality of Transformer. Secondly, we develop an Orthogonal Regularization to avoid learning redundant heads in the multi-head self-attention mechanism. Thirdly, we design a Trading Gap Splitter for Transformer to learn hierarchical features of high-frequency finance data. Compared with other popular recurrent neural networks such as LSTM, the proposed method has the advantage to mine extremely long-term dependencies from financial time series. Experimental results show our proposed models outperform several competitive methods in stock price prediction tasks for the NASDAQ exchange market and the China A-shares market.

Entity resolution (ER) aims to identify data records referring to the same real-world entity. Most existing ER approaches rely on the assumption that the entity records to be resolved are homogeneous, i.e., their attributes are aligned. Unfortunately, entities in real-world datasets are often heterogeneous, usually coming from different sources and being represented using different attributes. Furthermore, the entities’ attribute values may be redundant, noisy, missing, misplaced, or misspelled—we refer to it as the dirty data problem. To resolve the above problems, this paper proposes an end-to-end hierarchical matching network (HierMatcher) for entity resolution, which can jointly match entities in three levels—token, attribute, and entity. At the token level, a cross-attribute token alignment and comparison layer is designed to adaptively compare heterogeneous entities. At the attribute level, an attribute-aware attention mechanism is proposed to denoise dirty attribute values. Finally, the entity level matching layer effectively aggregates all matching evidence for the final ER decisions. Experimental results show that our method significantly outperforms previous ER methods on homogeneous, heterogeneous and dirty datasets.

Conceptual spaces are geometric meaning representations in which similar entities are represented by similar vectors. They are widely used in cognitive science, but there has been relatively little work on learning such representations from data. In particular, while standard representation learning methods can be used to induce vector space embeddings from text corpora, these differ from conceptual spaces in two crucial ways. First, the dimensions of a conceptual space correspond to salient semantic features, known as quality dimensions, whereas the dimensions of learned vector space embeddings typically lack any clear interpretation. This has been partially addressed in previous work, which has shown that it is possible to identify directions in learned vector spaces which capture semantic features. Second, conceptual spaces are normally organised into a set of domains, each of which is associated with a separate vector space. In contrast, learned embeddings represent all entities in a single vector space. Our hypothesis in this paper is that such single-space representations are sub-optimal for learning quality dimensions, due to the fact that semantic features are often only relevant to a subset of the entities. We show that this issue can be mitigated by identifying features in a hierarchical fashion. Intuitively, the top-level features split the vector space into different domains, making it possible to subsequently identify domain-specific quality dimensions.

2D image-based 3D shape retrieval has become a hot research topic since its wide industrial applications and academic significance. However, existing view-based 3D shape retrieval methods are restricted by two settings, 1) learn the common-class features while neglecting the instance visual characteristics, 2) narrow the global domain variations while ignoring the local semantic variations in each category. To overcome these problems, we propose a novel hierarchical instance feature alignment (HIFA) method for this task. HIFA consists of two modules, cross-modal instance feature learning and hierarchical instance feature alignment. Specifically, we first use CNN to extract both 2D image and multi-view features. Then, we maximize the mutual information between the input data and the high-level feature to preserve as much as visual characteristics of an individual instance. To mix up the features in two domains, we enforce feature alignment considering both global domain and local semantic levels. By narrowing the global domain variations we impose the identical large norm restriction on both 2D and 3D feature-norm expectations to facilitate more transferable possibility. By narrowing the local variations we propose to minimize the distance between two centroids of the same class from different domains to obtain semantic consistency. Extensive experiments on two popular and novel datasets, MI3DOR and MI3DOR-2, validate the superiority of HIFA for 2D image-based 3D shape retrieval task.

The task of Grounded Video Description~(GVD) is to generate sentences whose objects can be grounded with the bounding boxes in the video frames. Existing works often fail to exploit structural information both in modeling the relationships among the region proposals and in attending them for text generation. To address these issues, we cast the GVD task as a spatial-temporal Graph-to-Sequence learning problem, where we model video frames as spatial-temporal sequence graph in order to better capture implicit structural relationships. In particular, we exploit two ways to construct a sequence graph that captures spatial-temporal correlations among different objects in each frame and further present a novel graph topology refinement technique to discover optimal underlying graph structure. In addition, we also present hierarchical attention mechanism to attend sequence graph in different resolution levels for better generating the sentences. Our extensive experiments demonstrate the effectiveness of our proposed method compared to state-of-the-art methods.

Representation learning has offered a revolutionary learning paradigm for various AI domains. In this survey, we examine and review the problem of representation learning with the focus on heterogeneous networks, which consists of different types of vertices and relations. The goal of this problem is to automatically project objects, most commonly, vertices, in an input heterogeneous network into a latent embedding space such that both the structural and relational properties of the network can be encoded and preserved. The embeddings (representations) can be then used as the features to machine learning algorithms for addressing corresponding network tasks. To learn expressive embeddings, current research developments can fall into two major categories: shallow embedding learning and graph neural networks. After a thorough review of the existing literature, we identify several critical challenges that remain unaddressed and discuss future directions. Finally, we build the Heterogeneous Graph Benchmark to facilitate open research for this rapidly-developing topic.

Code mixing (or code switching) is a common phenomenon observed in social-media content generated by a linguistically diverse user-base. Studies show that in the Indian sub-continent, a substantial fraction of social media posts exhibit code switching. While the difficulties posed by code mixed documents to further downstream analyses are well-understood, lending visibility to code mixed documents under certain scenarios may have utility that has been previously overlooked. For instance, a document written in a mixture of multiple languages can be partially accessible to a wider audience; this could be particularly useful if a considerable fraction of the audience lacks fluency in one of the component languages. In this paper, we provide a systematic approach to sample code mixed documents leveraging a polyglot embedding based method that requires minimal supervision. In the context of the 2019 India-Pakistan conflict triggered by the Pulwama terror attack, we demonstrate an untapped potential of harnessing code mixing for human well-being: starting from an existing hostility diffusing hope speech classifier solely trained on English documents, code mixed documents are utilized to perform cross-lingual sampling and retrieve hope speech content written in a low-resource but widely used language - Romanized Hindi. Our proposed pipeline requires minimal supervision and holds promise in substantially reducing web moderation efforts. A further exploratory study on a new COVID-19 data set introduced in this paper demonstrates the generalizability of our cross-lingual sampling technique.

By virtue of their expressive power, neural networks (NNs) are well suited to fitting large, complex datasets, yet they are also known to produce similar predictions for points outside the training distribution. As such, they are, like humans, under the influence of the Black Swan theory: models tend to be extremely "surprised" by rare events, leading to potentially disastrous consequences, while justifying these same events in hindsight. To avoid this pitfall, we introduce DENN, an ensemble approach building a set of Diversely Extrapolated Neural Networks that fits the training data and is able to generalize more diversely when extrapolating to novel data points. This leads DENN to output highly uncertain predictions for unexpected inputs. We achieve this by adding a diversity term in the loss function used to train the model, computed at specific inputs. We first illustrate the usefulness of the method on a low-dimensional regression problem. Then, we show how the loss can be adapted to tackle anomaly detection during classification, as well as safe imitation learning problems.

Multiscale modeling has yielded immense success on various machine learning tasks. However, it has not been properly explored for the prominent task of information diffusion, which aims to understand how information propagates along users in online social networks. For a specific user, whether and when to adopt a piece of information propagated from another user is affected by complex interactions, and thus, is very challenging to model. Current state-of-the-art techniques invoke deep neural models with vector representations of users. In this paper, we present a Hierarchical Information Diffusion (HID) framework by integrating user representation learning and multiscale modeling. The proposed framework can be layered on top of all information diffusion techniques that leverage user representations, so as to boost the predictive power and learning efficiency of the original technique. Extensive experiments on three real-world datasets showcase the superiority of our method.

Stochastic video generation methods predict diverse videos based on observed frames, where the main challenge lies in modeling the complex future uncertainty and generating realistic frames. Numerous of Recurrent-VAE-based methods have achieved state-of-the-art results. However, on the one hand, the independence assumption of the variables of approximate posterior limits the inference performance. On the other hand, although these methods adopt skip connections between encoder and decoder to utilize multi-level features, they still produce blurry generation due to the spatial misalignment between encoder and decoder features at different time steps. In this paper, we propose a hierarchical recurrent VAE with a feature aligner, which can not only relax the independence assumption in typical VAE but also use a feature aligner to enable the decoder to obtain the aligned spatial information from the last observed frames. The proposed model is named Hierarchical Stochastic Video Generation network with Aligned Features, referred to as HAF-SVG. Experiments on Moving-MNIST, BAIR, and KTH datasets demonstrate that hierarchical structure is helpful for modeling more accurate future uncertainty, and the feature aligner is beneficial to generate realistic frames. Besides, the HAF-SVG exceeds SVG on both prediction accuracy and the quality of generated frames.

We present a conditional text generation framework that posits sentential expressions of possible causes and effects. This framework depends on two novel resources we develop in the course of this work: a very large-scale collection of English sentences expressing causal patterns (CausalBank); and a refinement over previous work on constructing large lexical causal knowledge graphs (Cause Effect Graph). Further, we extend prior work in lexically-constrained decoding to support disjunctive positive constraints. Human assessment confirms that our approach gives high-quality and diverse outputs. Finally, we use CausalBank to perform continued training of an encoder supporting a recent state-of-the-art model for causal reasoning, leading to a 3-point improvement on the COPA challenge set, with no change in model architecture.

We consider learning a convex combination of basis models, and present some new theoretical and empirical results that demonstrate the effectiveness of a greedy approach. Theoretically, we first consider whether we can use linear, instead of convex, combinations, and obtain generalization results similar to existing ones for learning from a convex hull. We obtain a negative result that even the linear hull of very simple basis functions can have unbounded capacity, and is thus prone to overfitting; on the other hand, convex hulls are still rich but have bounded capacities. Secondly, we obtain a generalization bound for a general class of Lipschitz loss functions. Empirically, we first discuss how a convex combination can be greedily learned with early stopping, and how a convex combination can be non-greedily learned when the number of basis models is known a priori. Our experiments suggest that the greedy scheme is competitive with or better than several baselines, including boosting and random forests. The greedy algorithm requires little effort in hyper-parameter tuning, and also seems able to adapt to the underlying complexity of the problem. Our code is available at https://github.com/tan1889/gce.

Sleep stage classification is essential for sleep assessment and disease diagnosis. However, how to effectively utilize brain spatial features and transition information among sleep stages continues to be challenging. In particular, owing to the limited knowledge of the human brain, predefining a suitable spatial brain connection structure for sleep stage classification remains an open question. In this paper, we propose a novel deep graph neural network, named GraphSleepNet, for automatic sleep stage classification. The main advantage of the GraphSleepNet is to adaptively learn the intrinsic connection among different electroencephalogram (EEG) channels, represented by an adjacency matrix, thereby best serving the spatial-temporal graph convolution network (ST-GCN) for sleep stage classification. Meanwhile, the ST-GCN consists of graph convolutions for extracting spatial features and temporal convolutions for capturing the transition rules among sleep stages. Experiments on the Montreal Archive of Sleep Studies (MASS) dataset demonstrate that the GraphSleepNet outperforms the state-of-the-art baselines.

Conversational machine comprehension (MC) has proven significantly more challenging compared to traditional MC since it requires better utilization of conversation history. However, most existing approaches do not effectively capture conversation history and thus have trouble handling questions involving coreference or ellipsis. Moreover, when reasoning over passage text, most of them simply treat it as a word sequence without exploring rich semantic relationships among words. In this paper, we first propose a simple yet effective graph structure learning technique to dynamically construct a question and conversation history aware context graph at each conversation turn. Then we propose a novel Recurrent Graph Neural Network, and based on that, we introduce a flow mechanism to model the temporal dependencies in a sequence of context graphs. The proposed GraphFlow model can effectively capture conversational flow in a dialog, and shows competitive performance compared to existing state-of-the-art methods on CoQA, QuAC and DoQA benchmarks. In addition, visualization experiments show that our proposed model can offer good interpretability for the reasoning process.

Neural-symbolic computing has now become the subject of interest of both academic and industry research laboratories. Graph Neural Networks (GNNs) have been widely used in relational and symbolic domains, with widespread application of GNNs in combinatorial optimization, constraint satisfaction, relational reasoning and other scientific domains. The need for improved explainability, interpretability and trust of AI systems in general demands principled methodologies, as suggested by neural-symbolic computing. In this paper, we review the state-of-the-art on the use of GNNs as a model of neural-symbolic computing. This includes the application of GNNs in several domains as well as their relationship to current developments in neural-symbolic computing.

Graph neural networks (GNNs) emerged recently as a powerful tool for analyzing non-Euclidean data such as social network data. Despite their success, the design of graph neural networks requires heavy manual work and domain knowledge. In this paper, we present a graph neural architecture search method (GraphNAS) that enables automatic design of the best graph neural architecture based on reinforcement learning. Specifically, GraphNAS uses a recurrent network to generate variable-length strings that describe the architectures of graph neural networks, and trains the recurrent network with policy gradient to maximize the expected accuracy of the generated architectures on a validation data set. Furthermore, to improve the search efficiency of GraphNAS on big networks, GraphNAS restricts the search space from an entire architecture space to a sequential concatenation of the best search results built on each single architecture layer. Experiments on real-world datasets demonstrate that GraphNAS can design a novel network architecture that rivals the best human-invented architecture in terms of validation set accuracy. Moreover, in a transfer learning task we observe that graph neural architectures designed by GraphNAS, when transferred to new datasets, still gain improvement in terms of prediction accuracy.

Gradient perturbation, widely used for differentially private optimization, injects noise at every iterative update to guarantee differential privacy. Previous work first determines the noise level that can satisfy the privacy requirement and then analyzes the utility of noisy gradient updates as in the non-private case. In contrast, we explore how the privacy noise affects the optimization property. We show that for differentially private convex optimization, the utility guarantee of differentially private (stochastic) gradient descent is determined by an expected curvature rather than the minimum curvature. The expected curvature, which represents the average curvature over the optimization path, is usually much larger than the minimum curvature. By using the expected curvature, we show that gradient perturbation can achieve a significantly improved utility guarantee that can theoretically justify the advantage of gradient perturbation over other perturbation methods. Finally, our extensive experiments suggest that gradient perturbation with the advanced composition method indeed outperforms other perturbation approaches by a large margin, matching our theoretical findings.

Goal recognition is the task of recognizing the objective of agents based on online observations of their behavior. Goal recognition design (GRD), the focus of this survey, facilitates goal recognition by the analysis and redesign of goal recognition models. In a nutshell, given a model of a domain and a set of possible goals, a solution to a GRD problem determines: (1) to what extent do actions performed by an agent reveal the agent’s objective? and (2) what is the best way to modify the model so that the objective of an agent can be detected as early as possible? GRD answers these questions by offering a solution for assessing and minimizing the maximal progress of any agent before recognition is guaranteed. This approach is relevant to any domain in which efficient goal recognition is essential and in which the model can be redesigned. Applications include intrusion detection, assisted cognition, computer games, and human-robot collaboration. This survey presents the solutions developed for evaluation and optimization in the GRD context, a discussion on the use of GRD in a variety of real-world applications, and suggestions of possible future avenues of GRD research.

Entity interaction prediction is essential in many important applications such as chemistry, biology, material science, and medical science. The problem becomes quite challenging when each entity is represented by a complex structure, namely structured entity, because two types of graphs are involved: local graphs for structured entities and a global graph to capture the interactions between structured entities. We observe that existing works on structured entity interaction prediction cannot properly exploit the unique graph of graphs model. In this paper, we propose a Graph of Graphs Neural Network, namely GoGNN, which extracts the features in both structured entity graphs and the entity interaction graph in a hierarchical way. We also propose the dual-attention mechanism that enables the model to preserve the neighbor importance in both levels of graphs. Extensive experiments on real-world datasets show that GoGNN outperforms the state-of-the-art methods on two representative structured entity interaction prediction tasks: chemical-chemical interaction prediction and drug-drug interaction prediction. Our code is available at Github.