Ke Ning

EGF-tree: an energy-efficient index tree for facilitating multi-region query aggregation in the internet of things

Spatial index trees constructed in wireless sensor networks are used to determine the sensors which can participate the query accurately and quickly. Most of index trees are constructed based on the parent--child node relation in network structure like routing tree, in which message sending for parent node selection will consume more energy. Due to energy being the important factor considered in wireless sensor networks, we design an energy-efficient index tree based on grid division and minimum energy merging principle in the skewness distribution of sensor nodes. Multi-region aggregation queries are carried on in our proposed index tree, which mainly focuses on region re-combination. Experimental results show that the energy consumption for multi-region aggregation queries are reduced compared to the original index tree.

Influence Analysis Based Expert Finding Model and Its Applications in Enterprise Social Network

In this paper, we propose a novel model for finding experts on a given topic using social influence analysis in enterprise social network. In enterprise social networks, employees usually talk about some topics relevant to their tasks. With the integration of social technology in BPM (Business Process Management), more expertise characteristics are reflected by their actions in social networks. Social networks became an important place for sharing expertise. We explore the potential of enterprise social networks, such as Yammer and IBM Connections, as a source of expertise evidence. In this work, we utilized influence analysis approach to find experts in enterprise social network. Generally, not all experts have the habit of sharing their expertise in social networks. So expert finding approaches, such as simply using link analysis, are of limited use. Our approach can address this problem. The experimental results show that the proposed approach can find real experts, not just managers with higher influence. Empirical results have also been presented to demonstrate the effectiveness of the proposed models.

A Sub-Chain Ranking and Recommendation Mechanism for Facilitating Geospatial Web Service Composition

With a huge volume of geospatial information being collected and a huge number of domain-specific functions being developed for processing these geospatial information, an increasing number of Open Geospatial Consortium Web services OWSs are built and being available on the Web for the accessibility and processing of these information. Given the specific requirement specified by a certain user, normally, a composition or chain of OWSs, rather than a single OWS, can fulfill this requirement. Consequently, retrieving and recommending sub-chains of possible service invocations is an important research challenge. Leveraging the semantic similarity between the name and text description of parameters, a degree that represents the invocation possibility between operations in OWSs is calculated. Thereafter, a service network model is constructed for capturing possible invocations between operations. Given a users requirement which is represented in terms of a pair of initial and ending operations, possible sub-chains of operations are retrieved, ranked and recommended. Based on which the user can select the most appropriate sub-chain with respect to her specific requirement. The result of evaluation leveraging a real OWSs set indicates that our technique is applicable in real applications from both the functional and performance perspectives.

Leverage Personal Cloud Storage Services to Provide Shared Storage for Team Collaboration

With the rapid development of cloud computing technology, cloud-based team collaboration applications are becoming popular on the Web. Among all the required features for a typical team collaboration application, shared storage for referred documents or produced artifacts by the team is a must-have one. However, existing shared storage solutions for team collaboration applications are far from satisfaction. Some of them rely on self-built storage infrastructure, which could be a big burden, especially for those small or medium vendors. With the prevalence of personal cloud storage services, such as Dropbox and Google Drive, more team collaboration applications allow users to share files from their personal cloud-storage spaces through external shared links, which can partly solve the problem. However, this method is not convenient for team collaboration, neither safe enough. This paper presents an approach to leverage third-part personal cloud-storage services to provide shared storage for team collaboration applications. Compared to existing approaches, our approach provides sophisticated mechanisms to make sure its more convenient and safer. It brings benefits in three folds: for users, it improves the utilization of personal cloud storage space, for vendors of personal cloud storage service, it helps attract users to use their services, for vendors of team collaboration applications, it reduces the burden of developing self-built storage infrastructure. The approach has been tested in kAct, a task-based team collaboration application provided by Kingdee, and the results are promising.

Parallel Matrix Multiplication Algorithm Based on Vector Linear Combination Using MapReduce

Matrix multiplication is used in a variety of applications. It requires a lot of computation time especially for large-scale matrices. Parallel processing is a good choice for matrix multiplication operation. To overcome the efficiencies of existing algorithms for parallel matrix multiplication, a matrix multiplication processing scheme based on vector linear combination (VLC) was presented. The VLC scheme splits the matrix multiplication procedure into two steps. The first step obtains the weighted vectors by scalar multiplication. The second step gets the final result through a linear combination of the weighted vectors with identical row numbers. We present parallel matrix multiplication implementations using MapReduce (MR) based on VLC scheme and explain in detail the MR job. The map method receives the matrix input and generates intermediate (key, value) pairs according to the VLC scheme requirement. The reduce method conducts the scalar multiplication and vectors summation. In the end, the reduce method outputs the result in the way of row vector. Then performance theoretical analysis and experiment result comparing with other algorithms are proposed. Algorithm presented in this paper needs less computation time than other algorithms. Finally, we conclude the paper and propose future works.

Geospatial Web Service Sub-chain Ranking and Recommendation

Nowadays, an increasing number of geospatial Web services (GWSs) are built and being available on the Web for the accessibility and processing of geospatial information. Given the requirement specified by certain users, normally a composition of GWSs, rather than a single GWS, can fulfill this requirement. Consequently, retrieving and recommending sub-chains of possible service invocations is an important research challenge. Leveraging the semantic similarity between the name and text description of parameters, the degree that represents the invocation possibility between operations in GWSs is calculated. The service network model is constructed to capture the possible invocations between operations. Given a users requirement which is represented in terms of the initial and ending operations, possible sub-chains of operations are retrieved, ranked and recommended. Thereafter, the user can select the most appropriate sub-chain with respect to her specific requirement. The evaluation result based on real GWSs set shows that our technique is applicable in real applications.

Cache-Based Periodic Query Optimization for Wireless Sensor Networks

Contiguous queries in wireless sensor networks may have some regions overlapping, and sensory data retrieved by recent queries may be used for answering the queries forthcoming, when these data are fresh enough. To support this query answering strategy, we propose a popularity-based caching mechanism for optimizing the periodic query processing. Specifically, the network region is divided using a cell-based manner, where each grid cell is abstracted as an elementary unit for the caching purpose. Fresh sensory data are cached in the memory of the sink node. The popularity of grid cells are calculated leveraging the queries conducted in recent time slots, which reflects the possibility that grid cells may be covered by the queries forthcoming. Prefetching may be performed for grid cells with a higher degree of popularity when their sensory data re missed in the cache. These cached sensory data are used for facilitating the query answering afterwards.

Services for Context Aware Knowledge Enhancement and Its Application in the Chinese Enterprise Management Tank (CEMT)

In the era of knowledge economy, knowledge resources have become the most valuable assets for enterprises. To better understand and reuse knowledge, it is necessary to relate it with the context in which the knowledge is generated and used. This is a process that usually occurs in an experienced knowledge-workers mind and without efficient supporting tools. This paper proposes an approach for the acquisition and utilization of context for the enhancement of knowledge and with a particular focus on methods to enable context extraction from industrial settings. The approach adopts a knowledge context ontology, to correlate knowledge and its context in the high-level activities of a knowledge worker. Knowledge context are extracted by utilizing a combination of methods including context identification, context reasoning, and context similarity measurement. Based on the proposed approach, a set of services for context aware knowledge enhancement are developed and applied in The Chinese Enterprise Management Tank (CEMT), a knowledge sharing and reusing platform for business management knowledge workers in all around China.

An Energy-Aware Service Composition Mechanism in Service-Oriented Wireless Sensor Networks

With the wide-adoption of the Internet of Things, heterogeneous smart things, serving as sensor nodes, require to work in a collective fashion for achieving complex applications. To address this challenge, this article proposes a service-oriented wireless sensor networks (WSNs) framework, where sensor nodes are encapsulated and represented as WSN services, which are energy-aware, and typically have constraints on their spatial and temporal aspects. Generally, WSN services are categorized into service classes according to the their functionalities. Consequently, service classes chains are generated with respect to the requirement of domain applications, and the composition of WSN services is constructed through discovering and selecting appropriate WSN services as the instantiation of service classes contained in chains. This WSN services composition is reduced to a multi-objective and multi-constrained optimization problem, which can be solved through adopting particle swarm optimization (PSO) algorithm and genetic algorithm (GA). Experimental evaluation shows that PSO outperforms GA in finding approximately optimal WSN services compositions.

Layer-Hierarchical Scientific Workflow Recommendation

This article proposes to identify and recommend scientific workflows to promote their reuse and repurposing. Specifically, a scientific workflow is converted into a layer hierarchy, which specifies hierarchical relations between this workflow, its sub-workflows, and activities. Semantic similarity is calculated between layer hierarchies of workflows in order to construct a scientific workflow network model. A graph-skeleton based clustering method is adopted for grouping layer hierarchies into clusters. Barycenters in clusters are identified for facilitating cluster identification and workflow ranking and recommendation. Experimental result shows that this technique is efficient and accurate on ranking and recommending appropriate clusters and scientific workflows.