Large transformer models have shown extraordinary success in achieving state-of-the-art results in many natural language processing applications. However, training and deploying these models can be prohibitively costly for long sequences, as the standard self-attention mechanism of the Transformer uses $O(n^2)$ time and space with respect to sequence length. In this paper, we demonstrate that the self-attention mechanism can be approximated by a low-rank matrix. We further exploit this finding to propose a new self-attention mechanism, which reduces the overall self-attention complexity from $O(n^2)$ to $O(n)$ in both time and space. The resulting linear transformer, the \textit{Linformer}, performs on par with standard Transformer models, while being much more memory- and time-efficient.
Recent work has suggested that language models (LMs) store both common-sense and factual knowledge learned from pre-training data. In this paper, we leverage this implicit knowledge to create an effective end-to-end fact checker using a solely a language model, without any external knowledge or explicit retrieval components. While previous work on extracting knowledge from LMs have focused on the task of open-domain question answering, to the best of our knowledge, this is the first work to examine the use of language models as fact checkers. In a closed-book setting, we show that our zero-shot LM approach outperforms a random baseline on the standard FEVER task, and that our fine-tuned LM compares favorably with standard baselines. Though we do not ultimately outperform methods which use explicit knowledge bases, we believe our exploration shows that this method is viable and has much room for exploration.
With several advantages and as an alternative to predict physics field, machine learning methods can be classified into two distinct types: data-driven relying on training data and physics-driven using physics law. Choosing heat conduction problem as an example, we compared the data- and physics-driven learning process with deep Convolutional Neural Networks (CNN). It shows that the convergences of the error to ground truth solution and the residual of heat conduction equation exhibit remarkable differences. Based on this observation, we propose a combined-driven method for learning acceleration and more accurate solutions. With a weighted loss function, reference data and physical equation are able to simultaneously drive the learning. Several numerical experiments are conducted to investigate the effectiveness of the combined method. For the data-driven based method, the introduction of physical equation not only is able to speed up the convergence, but also produces physically more consistent solutions. For the physics-driven based method, it is observed that the combined method is able to speed up the convergence up to 49.0\% by using a not very restrictive coarse reference.
As an important technology in 3D mapping, autonomous driving, and robot navigation, LiDAR odometry is still a challenging task. Appropriate data structure and unsupervised deep learning are the keys to achieve an easy adjusted LiDAR odometry solution with high performance. Utilizing compact 2D structured spherical ring projection model and voxel model which preserves the original shape of input data, we propose a fully unsupervised Convolutional Auto-Encoder based LiDAR Odometry (CAE-LO) that detects interest points from spherical ring data using 2D CAE and extracts features from multi-resolution voxel model using 3D CAE. We make several key contributions: 1) experiments based on KITTI dataset show that our interest points can capture more local details to improve the matching success rate on unstructured scenarios and our features outperform state-of-the-art by more than 50% in matching inlier ratio; 2) besides, we also propose a keyframe selection method based on matching pairs transferring, an odometry refinement method for keyframes based on extended interest points from spherical rings, and a backward pose update method. The odometry refinement experiments verify the proposed ideas' feasibility and effectiveness.
We present BlockBERT, a lightweight and efficient BERT model that is designed to better modeling long-distance dependencies. Our model extends BERT by introducing sparse block structures into the attention matrix to reduce both memory consumption and training time, which also enables attention heads to capture either short- or long-range contextual information. We conduct experiments on several benchmark question answering datasets with various paragraph lengths. Results show that BlockBERT uses 18.7-36.1% less memory and reduces the training time by 12.0-25.1%, while having comparable and sometimes better prediction accuracy, compared to an advanced BERT-based model, RoBERTa.
We study the problem of large-scale network embedding, which aims to learn latent representations for network mining applications. Previous research shows that 1) popular network embedding benchmarks, such as DeepWalk, are in essence implicitly factorizing a matrix with a closed form, and 2)the explicit factorization of such matrix generates more powerful embeddings than existing methods. However, directly constructing and factorizing this matrix---which is dense---is prohibitively expensive in terms of both time and space, making it not scalable for large networks. In this work, we present the algorithm of large-scale network embedding as sparse matrix factorization (NetSMF). NetSMF leverages theories from spectral sparsification to efficiently sparsify the aforementioned dense matrix, enabling significantly improved efficiency in embedding learning. The sparsified matrix is spectrally close to the original dense one with a theoretically bounded approximation error, which helps maintain the representation power of the learned embeddings. We conduct experiments on networks of various scales and types. Results show that among both popular benchmarks and factorization based methods, NetSMF is the only method that achieves both high efficiency and effectiveness. We show that NetSMF requires only 24 hours to generate effective embeddings for a large-scale academic collaboration network with tens of millions of nodes, while it would cost DeepWalk months and is computationally infeasible for the dense matrix factorization solution. The source code of NetSMF is publicly available (https://github.com/xptree/NetSMF).
Social and information networking activities such as on Facebook, Twitter, WeChat, and Weibo have become an indispensable part of our everyday life, where we can easily access friends' behaviors and are in turn influenced by them. Consequently, an effective social influence prediction for each user is critical for a variety of applications such as online recommendation and advertising. Conventional social influence prediction approaches typically design various hand-crafted rules to extract user- and network-specific features. However, their effectiveness heavily relies on the knowledge of domain experts. As a result, it is usually difficult to generalize them into different domains. Inspired by the recent success of deep neural networks in a wide range of computing applications, we design an end-to-end framework, DeepInf, to learn users' latent feature representation for predicting social influence. In general, DeepInf takes a user's local network as the input to a graph neural network for learning her latent social representation. We design strategies to incorporate both network structures and user-specific features into convolutional neural and attention networks. Extensive experiments on Open Academic Graph, Twitter, Weibo, and Digg, representing different types of social and information networks, demonstrate that the proposed end-to-end model, DeepInf, significantly outperforms traditional feature engineering-based approaches, suggesting the effectiveness of representation learning for social applications.
To enable efficient exploration of Web-scale scientific knowledge, it is necessary to organize scientific publications into a hierarchical concept structure. In this work, we present a large-scale system to (1) identify hundreds of thousands of scientific concepts, (2) tag these identified concepts to hundreds of millions of scientific publications by leveraging both text and graph structure, and (3) build a six-level concept hierarchy with a subsumption-based model. The system builds the most comprehensive cross-domain scientific concept ontology published to date, with more than 200 thousand concepts and over one million relationships.