arXiv subjects

Meihui Zhang

Publications and source records attributed to Meihui Zhang.

At least 37 records · Page 2Linked to original sources

arXiv2021

MLCask: Efficient Management of Component Evolution in Collaborative Data Analytics Pipelines

With the ever-increasing adoption of machine learning for data analytics, maintaining a machine learning pipeline is becoming more complex as both the datasets and trained models evolve with time. In a collaborative environment, the changes and updates due to pipeline evolution often cause cumbersome coordination and maintenance work, raising the costs and making it hard to use. Existing solutions, unfortunately, do not address the version evolution problem, especially in a collaborative environment where non-linear version control semantics are necessary to isolate operations made by different user roles. The lack of version control semantics also incurs unnecessary storage consumption and lowers efficiency due to data duplication and repeated data pre-processing, which are avoidable. In this paper, we identify two main challenges that arise during the deployment of machine learning pipelines, and address them with the design of versioning for an end-to-end analytics system MLCask. The system supports multiple user roles with the ability to perform Git-like branching and merging operations in the context of the machine learning pipelines. We define and accelerate the metric-driven merge operation by pruning the pipeline search tree using reusable history records and pipeline compatibility information. Further, we design and implement the prioritized pipeline search, which gives preference to the pipelines that probably yield better performance. The effectiveness of MLCask is evaluated through an extensive study over several real-world deployment cases. The performance evaluation shows that the proposed merge operation is up to 7.8x faster and saves up to 11.9x storage space than the baseline method that does not utilize history records.

cs.SE↗

arXiv2021

Blockchains vs. Distributed Databases: Dichotomy and Fusion

Blockchain has come a long way: a system that was initially proposed specifically for cryptocurrencies is now being adapted and adopted as a general-purpose transactional system. As blockchain evolves into another data management system, the natural question is how it compares against distributed database systems. Existing works on this comparison focus on high-level properties, such as security and throughput. They stop short of showing how the underlying design choices contribute to the overall differences. Our work fills this important gap and provides a principled framework for analyzing the emerging trend of blockchain-database fusion. We perform a twin study of blockchains and distributed database systems as two types of transactional systems. We propose a taxonomy that illustrates the dichotomy across four dimensions, namely replication, concurrency, storage, and sharding. Within each dimension, we discuss how the design choices are driven by two goals: security for blockchains, and performance for distributed databases. To expose the impact of different design choices on the overall performance, we conduct an in-depth performance analysis of two blockchains, namely Quorum and Hyperledger Fabric, and two distributed databases, namely TiDB, and etcd. Lastly, we propose a framework for back-of-the-envelope performance forecast of blockchain-database hybrids.

cs.DB↗

arXiv2020

Spitz: A Verifiable Database System

Databases in the past have helped businesses maintain and extract insights from their data. Today, it is common for a business to involve multiple independent, distrustful parties. This trend towards decentralization introduces a new and important requirement to databases: the integrity of the data, the history, and the execution must be protected. In other words, there is a need for a new class of database systems whose integrity can be verified (or verifiable databases). In this paper, we identify the requirements and the design challenges of verifiable databases.We observe that the main challenges come from the need to balance data immutability, tamper evidence, and performance. We first consider approaches that extend existing OLTP and OLAP systems with support for verification. We next examine a clean-slate approach, by describing a new system, Spitz, specifically designed for efficiently supporting immutable and tamper-evident transaction management. We conduct a preliminary performance study of both approaches against a baseline system, and provide insights on their performance.

cs.DB↗

arXiv2020

Effective and Efficient Dropout for Deep Convolutional Neural Networks

Convolutional Neural networks (CNNs) based applications have become ubiquitous, where proper regularization is greatly needed. To prevent large neural network models from overfitting, dropout has been widely used as an efficient regularization technique in practice. However, many recent works show that the standard dropout is ineffective or even detrimental to the training of CNNs. In this paper, we revisit this issue and examine various dropout variants in an attempt to improve existing dropout-based regularization techniques for CNNs. We attribute the failure of standard dropout to the conflict between the stochasticity of dropout and its following Batch Normalization (BN), and propose to reduce the conflict by placing dropout operations right before the convolutional operation instead of BN, or totally address this issue by replacing BN with Group Normalization (GN). We further introduce a structurally more suited dropout variant Drop-Conv2d, which provides more efficient and effective regularization for deep CNNs. These dropout variants can be readily integrated into the building blocks of CNNs and implemented in existing deep learning platforms. Extensive experiments on benchmark datasets including CIFAR, SVHN and ImageNet are conducted to compare the existing building blocks and the proposed ones with dropout training. Results show that our building blocks improve over state-of-the-art CNNs significantly, which is mainly due to the better regularization and implicit model ensemble effect.

cs.LG↗

arXiv2020

ForkBase: Immutable, Tamper-evident Storage Substrate for Branchable Applications

Data collaboration activities typically require systematic or protocol-based coordination to be scalable. Git, an effective enabler for collaborative coding, has been attested for its success in countless projects around the world. Hence, applying the Git philosophy to general data collaboration beyond coding is motivating. We call it Git for data. However, the original Git design handles data at the file granule, which is considered too coarse-grained for many database applications. We argue that Git for data should be co-designed with database systems. To this end, we developed ForkBase to make Git for data practical. ForkBase is a distributed, immutable storage system designed for data version management and data collaborative operation. In this demonstration, we show how ForkBase can greatly facilitate collaborative data management and how its novel data deduplication technique can improve storage efficiency for archiving massive data versions.

cs.DB↗

arXiv2020

A Transactional Perspective on Execute-order-validate Blockchains

Smart contracts have enabled blockchain systems to evolve from simple cryptocurrency platforms, such as Bitcoin, to general transactional systems, such as Ethereum. Catering for emerging business requirements, a new architecture called execute-order-validate has been proposed in Hyperledger Fabric to support parallel transactions and improve the blockchain's throughput. However, this new architecture might render many invalid transactions when serializing them. This problem is further exaggerated as the block formation rate is inherently limited due to other factors beside data processing, such as cryptography and consensus. In this work, we propose a novel method to enhance the execute-order-validate architecture, by reducing invalid transactions to improve the throughput of blockchains. Our method is inspired by state-of-the-art optimistic concurrency control techniques in modern database systems. In contrast to existing blockchains that adopt database's preventive approaches which might abort serializable transactions, our method is theoretically more fine-grained. Specifically, unserializable transactions are aborted before ordering and the remaining transactions are guaranteed to be serializable. For evaluation, we implement our method in two blockchains respectively, FabricSharp on top of Hyperledger Fabric, and FastFabricSharp on top of FastFabric. We compare the performance of FabricSharp with vanilla Fabric and three related systems, two of which are respectively implemented with one standard and one state-of-the-art concurrency control techniques from databases. The results demonstrate that FabricSharp achieves 25% higher throughput compared to the other systems in nearly all experimental scenarios. Moreover, the FastFabricSharp's improvement over FastFabric is up to 66%.

cs.DC↗

arXiv2020

Analysis of Indexing Structures for Immutable Data

In emerging applications such as blockchains and collaborative data analytics, there are strong demands for data immutability, multi-version accesses, and tamper-evident controls. This leads to three new index structures for immutable data, namely Merkle Patricia Trie (MPT), Merkle Bucket Tree (MBT), and Pattern-Oriented-Split Tree (POS-Tree). Although these structures have been adopted in real applications, there is no systematic evaluation of their pros and cons in the literature. This makes it difficult for practitioners to choose the right index structure for their applications, as there is only a limited understanding of the characteristics of each index. To alleviate the above deficiency, we present a comprehensive analysis of the existing index structures for immutable data, evaluating both their asymptotic and empirical performance. Specifically, we show that MPT, MBT, and POS-Tree are all instances of a recently proposed framework, dubbed \my{Structurally Invariant and Reusable Indexes (SIRI)}. We propose to evaluate the SIRI instances based on five essential metrics: their efficiency for four index operations (i.e., lookup, update, comparison, and merge), as well as their \my{deduplication ratios} (i.e., the size of the index with deduplication over the size without deduplication). We establish the worst-case guarantees of each index in terms of these five metrics, and we experimentally evaluate all indexes in a large variety of settings. Based on our theoretical and empirical analysis, we conclude that POS-Tree is a favorable choice for indexing immutable data.

cs.DB↗

arXiv2020

Database Meets Deep Learning: Challenges and Opportunities

Deep learning has recently become very popular on account of its incredible success in many complex data-driven applications, such as image classification and speech recognition. The database community has worked on data-driven applications for many years, and therefore should be playing a lead role in supporting this new wave. However, databases and deep learning are different in terms of both techniques and applications. In this paper, we discuss research problems at the intersection of the two fields. In particular, we discuss possible improvements for deep learning systems from a database perspective, and analyze database applications that may benefit from deep learning techniques.

cs.DB↗

arXiv2019

The Disruptions of 5G on Data-driven Technologies and Applications

With 5G on the verge of being adopted as the next mobile network, there is a need to analyze its impact on the landscape of computing and data management. In this paper, we analyze the impact of 5G on both traditional and emerging technologies and project our view on future research challenges and opportunities. With a predicted increase of 10-100x in bandwidth and 5-10x decrease in latency, 5G is expected to be the main enabler for smart cities, smart IoT and efficient healthcare, where machine learning is conducted at the edge. In this context, we investigate how 5G can help the development of federated learning. Network slicing, another key feature of 5G, allows running multiple isolated networks on the same physical infrastructure. However, security remains the main concern in the context of virtualization, multi-tenancy and high device density. Formal verification of 5G networks can be applied to detect security issues in massive virtualized environments. In summary, 5G will make the world even more densely and closely connected. What we have experienced in 4G connectivity will pale in comparison to the vast amounts of possibilities engendered by 5G.

cs.NI↗

arXiv2019

5G: Agent for Further Digital Disruptive Transformations

The fifth-generation (5G) mobile communication technologies are on the way to be adopted as the next standard for mobile networking. It is therefore timely to analyze the impact of 5G on the landscape of computing, in particular, data management and data-driven technologies. With a predicted increase of 10-100$\times$ in bandwidth and 5-10$\times$ decrease in latency, 5G is expected to be the main enabler for edge computing which includes accessing cloud-like services, as well as conducting machine learning at the edge. In this paper, we examine the impact of 5G on both traditional and emerging technologies, and discuss research challenges and opportunities.

cs.NI↗

arXiv2018

Object-Level Representation Learning for Few-Shot Image Classification

Few-shot learning that trains image classifiers over few labeled examples per category is a challenging task. In this paper, we propose to exploit an additional big dataset with different categories to improve the accuracy of few-shot learning over our target dataset. Our approach is based on the observation that images can be decomposed into objects, which may appear in images from both the additional dataset and our target dataset. We use the object-level relation learned from the additional dataset to infer the similarity of images in our target dataset with unseen categories. Nearest neighbor search is applied to do image classification, which is a non-parametric model and thus does not need fine-tuning. We evaluate our algorithm on two popular datasets, namely Omniglot and MiniImagenet. We obtain 8.5\% and 2.7\% absolute improvements for 5-way 1-shot and 5-way 5-shot experiments on MiniImagenet, respectively. Source code will be published upon acceptance.

cs.CV↗

arXiv2018

PANDA: Facilitating Usable AI Development

Recent advances in artificial intelligence (AI) and machine learning have created a general perception that AI could be used to solve complex problems, and in some situations over-hyped as a tool that can be so easily used. Unfortunately, the barrier to realization of mass adoption of AI on various business domains is too high because most domain experts have no background in AI. Developing AI applications involves multiple phases, namely data preparation, application modeling, and product deployment. The effort of AI research has been spent mostly on new AI models (in the model training stage) to improve the performance of benchmark tasks such as image recognition. Many other factors such as usability, efficiency and security of AI have not been well addressed, and therefore form a barrier to democratizing AI. Further, for many real world applications such as healthcare and autonomous driving, learning via huge amounts of possibility exploration is not feasible since humans are involved. In many complex applications such as healthcare, subject matter experts (e.g. Clinicians) are the ones who appreciate the importance of features that affect health, and their knowledge together with existing knowledge bases are critical to the end results. In this paper, we take a new perspective on developing AI solutions, and present a solution for making AI usable. We hope that this resolution will enable all subject matter experts (eg. Clinicians) to exploit AI like data scientists.

cs.AI↗

arXiv2018

Rafiki: Machine Learning as an Analytics Service System

Big data analytics is gaining massive momentum in the last few years. Applying machine learning models to big data has become an implicit requirement or an expectation for most analysis tasks, especially on high-stakes applications.Typical applications include sentiment analysis against reviews for analyzing on-line products, image classification in food logging applications for monitoring user's daily intake and stock movement prediction. Extending traditional database systems to support the above analysis is intriguing but challenging. First, it is almost impossible to implement all machine learning models in the database engines. Second, expertise knowledge is required to optimize the training and inference procedures in terms of efficiency and effectiveness, which imposes heavy burden on the system users. In this paper, we develop and present a system, called Rafiki, to provide the training and inference service of machine learning models, and facilitate complex analytics on top of cloud platforms. Rafiki provides distributed hyper-parameter tuning for the training service, and online ensemble modeling for the inference service which trades off between latency and accuracy. Experimental results confirm the efficiency, effectiveness, scalability and usability of Rafiki.

cs.DB↗

arXiv2018

ForkBase: An Efficient Storage Engine for Blockchain and Forkable Applications

Existing data storage systems offer a wide range of functionalities to accommodate an equally diverse range of applications. However, new classes of applications have emerged, e.g., blockchain and collaborative analytics, featuring data versioning, fork semantics, tamper-evidence or any combination thereof. They present new opportunities for storage systems to efficiently support such applications by embedding the above requirements into the storage. In this paper, we present ForkBase, a storage engine specifically designed to provide efficient support for blockchain and forkable applications. By integrating the core application properties into the storage, ForkBase not only delivers high performance but also reduces development effort. Data in ForkBase is multi-versioned, and each version uniquely identifies the data content and its history. Two variants of fork semantics are supported in ForkBase to facilitate any collaboration workflows. A novel index structure is introduced to efficiently identify and eliminate duplicate content across data objects. Consequently, ForkBase is not only efficient in performance, but also in space requirement. We demonstrate the performance of ForkBase using three applications: a blockchain platform, a wiki engine and a collaborative analytics application. We conduct extensive experimental evaluation of these applications against respective state-of-the-art system. The results show that ForkBase achieves superior performance while significantly lowering the development cost.

cs.DB↗

arXiv2017

Cross-Domain Image Retrieval with Attention Modeling

With the proliferation of e-commerce websites and the ubiquitousness of smart phones, cross-domain image retrieval using images taken by smart phones as queries to search products on e-commerce websites is emerging as a popular application. One challenge of this task is to locate the attention of both the query and database images. In particular, database images, e.g. of fashion products, on e-commerce websites are typically displayed with other accessories, and the images taken by users contain noisy background and large variations in orientation and lighting. Consequently, their attention is difficult to locate. In this paper, we exploit the rich tag information available on the e-commerce websites to locate the attention of database images. For query images, we use each candidate image in the database as the context to locate the query attention. Novel deep convolutional neural network architectures, namely TagYNet and CtxYNet, are proposed to learn the attention weights and then extract effective representations of the images. Experimental results on public datasets confirm that our approaches have significant improvement over the existing methods in terms of the retrieval accuracy and efficiency.

cs.MM↗

arXiv2017

Untangling Blockchain: A Data Processing View of Blockchain Systems

Blockchain technologies are gaining massive momentum in the last few years. Blockchains are distributed ledgers that enable parties who do not fully trust each other to maintain a set of global states. The parties agree on the existence, values and histories of the states. As the technology landscape is expanding rapidly, it is both important and challenging to have a firm grasp of what the core technologies have to offer, especially with respect to their data processing capabilities. In this paper, we first survey the state of the art, focusing on private blockchains (in which parties are authenticated). We analyze both in-production and research systems in four dimensions: distributed ledger, cryptography, consensus protocol and smart contract. We then present BLOCKBENCH, a benchmarking framework for understanding performance of private blockchains against data processing workloads. We conduct a comprehensive evaluation of three major blockchain systems based on BLOCKBENCH, namely Ethereum, Parity and Hyperledger Fabric. The results demonstrate several trade-offs in the design space, as well as big performance gaps between blockchain and database systems. Drawing from design principles of database systems, we discuss several research directions for bringing blockchain performance closer to the realm of databases.

cs.DB↗

arXiv2017

Scaling Distributed Transaction Processing and Recovery based on Dependency Logging

DGCC protocol has been shown to achieve good performance on multi-core in-memory system. However, distributed transactions complicate the dependency resolution, and therefore, an effective transaction partitioning strategy is essential to reduce expensive multi-node distributed transactions. During failure recovery, log must be examined from the last checkpoint onwards and the affected transactions are re-executed based on the way they are partitioned and executed. Existing approaches treat both transaction management and recovery as two separate problems, even though recovery is dependent on the sequence in which transactions are executed. In this paper, we propose to treat the transaction management and recovery problems as one. We first propose an efficient Distributed Dependency Graph based Concurrency Control (DistDGCC) protocol for handling transactions spanning multiple nodes, and propose a new novel and efficient logging protocol called Dependency Logging that also makes use of dependency graphs for efficient logging and recovery. DistDGCC optimizes the average cost for each distributed transaction by processing transactions in batch. Moreover, it also reduces the effects of thread blocking caused by distributed transactions and consequently improves the runtime performance. Further, dependency logging exploits the same data structure that is used by DistDGCC to reduce the logging overhead, as well as the logical dependency information to improve the recovery parallelism. Extensive experiments are conducted to evaluate the performance of our proposed technique against state-of-the-art techniques. Experimental results show that DistDGCC is efficient and scalable, and dependency logging supports fast recovery with marginal runtime overhead. Hence, the overall system performance is significantly improved as a result.

cs.DB↗

arXiv2017

UStore: A Distributed Storage With Rich Semantics

Today's storage systems expose abstractions which are either too low-level (e.g., key-value store, raw-block store) that they require developers to re-invent the wheels, or too high-level (e.g., relational databases, Git) that they lack generality to support many classes of applications. In this work, we propose and implement a general distributed data storage system, called UStore, which has rich semantics. UStore delivers three key properties, namely immutability, sharing and security, which unify and add values to many classes of today's applications, and which also open the door for new applications. By keeping the core properties within the storage, UStore helps reduce application development efforts while offering high performance at hand. The storage embraces current hardware trends as key enablers. It is built around a data-structure similar to that of Git, a popular source code versioning system, but it also synthesizes many designs from distributed systems and databases. Our current implementation of UStore has better performance than general in-memory key-value storage systems, especially for version scan operations. We port and evaluate four applications on top of UStore: a Git-like application, a collaborative data science application, a transaction management application, and a blockchain application. We demonstrate that UStore enables faster development and the UStore-backed applications can have better performance than the existing implementations.

cs.DB↗