NSF Launches Large-Scale Network for Small-Scale Science

Label propagation spreads the soft labels from few labeled data to a large amount of unlabeled data according to the intrinsic graph structure. Nonetheless, most label propagation solutions work under relatively small-scale data and fail to cope with many real applications, such as social network analysis, where graphs usually have millions of nodes. In this paper, we propose a novel algorithm named \algo to deal with large-scale data. A lightweight iterative process derived from the well-known stochastic gradient descent strategy is used to reduce memory overhead and accelerate the solving process. We also give a theoretical analysis on the necessity of the warm-start technique for label propagation. Experiments show that our algorithm can handle million-scale graphs in few seconds while achieving highly competitive performance with existing algorithms.

Download Full-text

A Hierarchical and Abstraction-Based Blockchain Model

Applied Sciences ◽

10.3390/app9112343 ◽

2019 ◽

Vol 9 (11) ◽

pp. 2343 ◽

Cited By ~ 3

Author(s):

Swagatika Sahoo ◽

Akshay M. Fajge ◽

Raju Halder ◽

Agostino Cortesi

Keyword(s):

Large Scale ◽

Abstract Interpretation ◽

Organizational Structures ◽

Network Size ◽

Small Scale ◽

Time Interval ◽

Huge Number ◽

Large Scale Network ◽

Regular Time ◽

Scale Network

In the nine years since its launch, amid intense research, scalability is always a serious concern in blockchain, especially in case of large-scale network generating huge number of transaction-records. In this paper, we propose a hierarchical blockchain model characterized by: (1) each level maintains multiple local blockchain networks, (2) each local blockchain records local transactional activities, and (3) partial views (tunable w.r.t. precision) of different subsets of local blockchain-records are maintained in the blockchains at next level of the hierarchy. To meet this objective, we apply abstractions on a set of transaction-records in a regular time interval by following the Abstract Interpretation framework, which provides a tunable precision in various abstract domain and guarantees the soundness of the system. While this model suitably fits to the real-worlds organizational structures, the proposal is powerful enough to scale when large number of nodes participate in a network resulting into an enormous growth of the network-size and the number of transaction-records. We discuss experimental results on a small-scale network with three sub networks at lower-level and by abstracting the transaction-records in the abstract domain of intervals. The results are encouraging and clearly indicate the effectiveness of this approach to control exponential growth of blockchain size w.r.t. the total number of participants in the network.

Download Full-text

A Topology Visualization Early Warning Distribution Algorithm for Large-Scale Network Security Incidents

The Scientific World JOURNAL ◽

10.1155/2013/827376 ◽

2013 ◽

Vol 2013 ◽

pp. 1-7

Author(s):

Hui He ◽

Guotao Fan ◽

Jianwei Ye ◽

Weizhe Zhang

Keyword(s):

Network Security ◽

Network Topology ◽

Early Warning ◽

Early Warning System ◽

Large Scale ◽

Warning System ◽

Small Scale ◽

Large Scale Network ◽

Scale Network ◽

Security Incidents

It is of great significance to research the early warning system for large-scale network security incidents. It can improve the network system’s emergency response capabilities, alleviate the cyber attacks’ damage, and strengthen the system’s counterattack ability. A comprehensive early warning system is presented in this paper, which combines active measurement and anomaly detection. The key visualization algorithm and technology of the system are mainly discussed. The large-scale network system’s plane visualization is realized based on the divide and conquer thought. First, the topology of the large-scale network is divided into some small-scale networks by the MLkP/CR algorithm. Second, the sub graph plane visualization algorithm is applied to each small-scale network. Finally, the small-scale networks’ topologies are combined into a topology based on the automatic distribution algorithm of force analysis. As the algorithm transforms the large-scale network topology plane visualization problem into a series of small-scale network topology plane visualization and distribution problems, it has higher parallelism and is able to handle the display of ultra-large-scale network topology.

Download Full-text