Alibaba Group

About: Alibaba Group is a company organization based out in Hangzhou, China. It is known for research contribution in the topics: Computer science & Terminal (electronics). The organization has 6810 authors who have published 7389 publications receiving 55653 citations. The organization is also known as: Alibaba Group Holding Limited & Alibaba Group (Cayman Islands).

Interaction Embeddings for Prediction and Explanation in Knowledge Graphs

Abstract: Knowledge graph embedding aims to learn distributed representations for entities and relations, and is proven to be effective in many applications. Crossover interactions -- bi-directional effects between entities and relations --- help select related information when predicting a new triple, but haven't been formally discussed before. In this paper, we propose CrossE, a novel knowledge graph embedding which explicitly simulates crossover interactions. It not only learns one general embedding for each entity and relation as most previous methods do, but also generates multiple triple specific embeddings for both of them, named interaction embeddings. We evaluate embeddings on typical link prediction tasks and find that CrossE achieves state-of-the-art results on complex and more challenging datasets. Furthermore, we evaluate embeddings from a new perspective -- giving explanations for predicted triples, which is important for real applications. In this work, an explanation for a triple is regarded as a reliable closed-path between the head and the tail entity. Compared to other baselines, we show experimentally that CrossE, benefiting from interaction embeddings, is more capable of generating reliable explanations to support its predictions.

Combining Graph-Based Learning With Automated Data Collection for Code Vulnerability Detection

Abstract: This paper presents FUNDED (Flow-sensitive vUl-Nerability coDE Detection), a novel learning framework for building vulnerability detection models. Funded leverages the advances in graph neural networks (GNNs) to develop a novel graph-based learning method to capture and reason about the program’s control, data, and call dependencies. Unlike prior work that treats the program as a sequential sequence or an untyped graph, Funded learns and operates on a graph representation of the program source code, in which individual statements are connected to other statements through relational edges. By capturing the program syntax, semantics and flows, Funded finds better code representation for the downstream software vulnerability detection task. To provide sufficient training data to build an effective deep learning model, we combine probabilistic learning and statistical assessments to automatically gather high-quality training samples from open-source projects. This provides many real-life vulnerable code training samples to complement the limited vulnerable code samples available in standard vulnerability databases. We apply Funded to identify software vulnerabilities at the function level from program source code. We evaluate Funded on large real-world datasets with programs written in C, Java, Swift and Php, and compare it against six state-of-the-art code vulnerability detection models. Experimental results show that Funded significantly outperforms alternative approaches across evaluation settings.

Learning deep semantic segmentation network under multiple weakly-supervised constraints for cross-domain remote sensing image semantic segmentation

Abstract: Due to its wide applications, remote sensing (RS) image semantic segmentation has attracted increasing research interest in recent years. Benefiting from its hierarchical abstract ability, the deep semantic segmentation network (DSSN) has achieved tremendous success on RS image semantic segmentation and has gradually become the mainstream technology. However, the superior performance of DSSN highly depends on two conditions: (I) massive quantities of labeled training data exist; (II) the testing data seriously resemble the training data. In actual RS applications, it is difficult to fully meet these conditions due to the RS sensor variation and the distinct landscape variation in different geographic locations. To make DSSN fit the actual RS scenario, this paper exploits the cross-domain RS image semantic segmentation task, which means that DSSN is trained on one labeled dataset (i.e., the source domain) but is tested on another varied dataset (i.e., the target domain). In this setting, the performance of DSSN is inevitably very limited due to the data shift between the source and target domains. To reduce the disadvantageous influence of data shift, this paper proposes a novel objective function with multiple weakly-supervised constraints to learn DSSN for cross-domain RS image semantic segmentation. Through carefully examining the characteristics of cross-domain RS image semantic segmentation, multiple weakly-supervised constraints include the weakly-supervised transfer invariant constraint (WTIC), weakly-supervised pseudo-label constraint (WPLC) and weakly-supervised rotation consistency constraint (WRCC). Specifically, DualGAN is recommended to conduct unsupervised style transfer between the source and target domains to carry out WTIC. To make full use of the merits of multiple constraints, this paper presents a dynamic optimization strategy that dynamically adjusts the constraint weights of the objective function during the training process. With full consideration of the characteristics of the cross-domain RS image semantic segmentation task, this paper gives two cross-domain RS image semantic segmentation settings: (I) variation in geographic location and (II) variation in both geographic location and imaging mode. Extensive experiments demonstrate that our proposed method remarkably outperforms the state-of-the-art methods under both of these settings. The collected datasets and evaluation benchmarks have been made publicly available online ( https://github.com/te-shi/MUCSS ).

Unpaired Image Captioning via Scene Graph Alignments

Abstract: Most of current image captioning models heavily rely on paired image-caption datasets. However, getting large scale image-caption paired data is labor-intensive and time-consuming. In this paper, we present a scene graph-based approach for unpaired image captioning. Our framework comprises an image scene graph generator, a sentence scene graph generator, a scene graph encoder, and a sentence decoder. Specifically, we first train the scene graph encoder and the sentence decoder on the text modality. To align the scene graphs between images and sentences, we propose an unsupervised feature alignment method that maps the scene graph features from the image to the sentence modality. Experimental results show that our proposed model can generate quite promising results without using any image-caption training pairs, outperforming existing methods by a wide margin.

MLPerf Inference Benchmark

Abstract: Machine-learning (ML) hardware and software system demand is burgeoning. Driven by ML applications, the number of different ML inference systems has exploded. Over 100 organizations are building ML inference chips, and the systems that incorporate existing models span at least three orders of magnitude in power consumption and five orders of magnitude in performance; they range from embedded devices to data-center solutions. Fueling the hardware are a dozen or more software frameworks and libraries. The myriad combinations of ML hardware and ML software make assessing ML-system performance in an architecture-neutral, representative, and reproducible manner challenging. There is a clear need for industry-wide standard ML benchmarking and evaluation criteria. MLPerf Inference answers that call. In this paper, we present our benchmarking method for evaluating ML inference systems. Driven by more than 30 organizations as well as more than 200 ML engineers and practitioners, MLPerf prescribes a set of rules and best practices to ensure comparability across systems with wildly differing architectures. The first call for submissions garnered more than 600 reproducible inference-performance measurements from 14 organizations, representing over 30 systems that showcase a wide range of capabilities. The submissions attest to the benchmark's flexibility and adaptability.