Lan Huang

Link Clustering with Extended Link Similarity and EQ Evaluation Division.

Link Clustering (LC) is a relatively new method for detecting overlapping communities in networks. The basic principle of LC is to derive a transform matrix whose elements are composed of the link similarity of neighbor links based on the Jaccard distance calculation; then it applies hierarchical clustering to the transform matrix and uses a measure of partition density on the resulting dendrogram to determine the cut level for best community detection. However, the original link clustering method does not consider the link similarity of non-neighbor links, and the partition density tends to divide the communities into many small communities. In this paper, an Extended Link Clustering method (ELC) for overlapping community detection is proposed. The improved method employs a new link similarity, Extended Link Similarity (ELS), to produce a denser transform matrix, and uses the maximum value of EQ (an extended measure of quality of modularity) as a means to optimally cut the dendrogram for better partitioning of the original network space. Since ELS uses more link information, the resulting transform matrix provides a superior basis for clustering and analysis. Further, using the EQ value to find the best level for the hierarchical clustering dendrogram division, we obtain communities that are more sensible and reasonable than the ones obtained by the partition density evaluation. Experimentation on five real-world networks and artificially-generated networks shows that the ELC method achieves higher EQ and In-group Proportion (IGP) values. Additionally, communities are more realistic than those generated by either of the original LC method or the classical CPM method.

A link density clustering algorithm based on automatically selecting density peaks for overlapping community detection

In this paper, we proposed a link density clustering (LDC) method for overlapping community detection based on density peaks. We firstly use an extended cosine link distance metric to reflect the relationship of links. Then we introduce a clustering algorithm with fast search for solving the link clustering (LC) problem by density peaks with box plot strategy to determine the cluster centers automatically. Finally, we acquire both the link communities and the node communities. Our algorithm is compared with other representative algorithms through substantial experiments on real-world networks. The experimental results show that our algorithm consistently outperforms other algorithms in terms of modularity and coverage.

Evaluation of scientific publications with hesitant fuzzy uncertain linguistic and semantic information

Scientific and technological papers play a fundamental role in the scientific and technological innovation of countries. The quality control of scientific and technological articles is vital to the journals and management of personnel. This paper investigates multiple attribute decision-making problems with the application of hesitant fuzzy uncertain linguistic information. Motivated by the ideal traditional I-COA operator, an induced hesitant fuzzy uncertain linguistic correlated averaging (IHFULCA) operator is developed. The IHFULCA operator was used to develop approaches to solve hesitant fuzzy uncertain linguistic multiple attribute decision-making problems. Finally, a practical example for evaluating the academic value of scientific and technological papers is provided to verify the developed approach and demonstrate its practicality and effectiveness.

e-Distance Weighted Support Vector Regression

We propose a novel support vector regression approach called (varepsilon )-Distance Weighted Support Vector Regression ((varepsilon )-DWSVR). (varepsilon )-DWSVR specifically addresses a challenging issue in support vector regression: how to deal with the situation when the distribution of the internal data in the (varepsilon )-tube is different from that of the boundary data containing support vectors. The proposed (varepsilon )-DWSVR optimizes the minimum margin and the mean of functional margin simultaneously to tackle this issue. To solve the new optimization problem arising from (varepsilon )-DWSVR, we adopt dual coordinate descent (DCD) with kernel functions for medium-scale problems and also employ averaged stochastic gradient descent (ASGD) to make (varepsilon )-DWSVR scalable to larger problems. We report promising results obtained by (varepsilon )-DWSVR in comparison with five popular regression methods on sixteen UCI benchmark datasets.

A Novel Spatio-Temporal Data Storage and Index Method for ARM-Based Hadoop Server

During the past decade, a vast number of GPS devices have produced massive amounts of data containing both time and spatial information. This poses a great challenge for traditional spatial databases. With the development of distributed cloud computing, many high-performance cloud platforms have been built, which can be used to process such spatio-temporal data. In this research, to store and process data in an effective and green way, we propose the following solutions: firstly, we build a Hadoop cloud computing platform using Cubieboards2, an ARM development board with A20 processors; secondly, we design two types of indexes for different types of spatio-temporal data at the HDFS level. We use a specific partitioning strategy to divide data in order to ensure load balancing and efficient range query. To improve the efficiency of disk utilisation and network transmission, we also optimise the storage structure. The experimental results show that our cloud platform is highly scalable, and the two types of indexes are effective for spatio-temporal data storage optimisation and they can help achieve high retrieval efficiency.

Link community detection based on line graphs with a novel link similarity measure

Link community gradually unfolds its capacity in complex network research. In this paper, a novel link similarity measure on line graphs is proposed. This measure can be adapted to different types of networks with an adjustable parameter. We prove its value converges to a limit on line graphs with the relationship of the nonneighbor links taken into account. Based on this similarity measure, we propose a novel link community detection algorithm for link clustering on line graphs. The detection algorithm combines the novel link similarity measure with the classic Markov Cluster (MCL) Algorithm and determines the link community partitions by calculating an extended modularity measure. Extensive experiments on two types of complex networks demonstrate the effectiveness, reliability and rationality of our solution in contrast to the other two classical algorithms.

A Central Edge Selection Based Overlapping Community Detection Algorithm for the Detection of Overlapping Structures in Protein–Protein Interaction Networks

Overlapping structures of protein–protein interaction networks are very prevalent in different biological processes, which reflect the sharing mechanism to common functional components. The overlapping community detection (OCD) algorithm based on central node selection (CNS) is a traditional and acceptable algorithm for OCD in networks. The main content of CNS is the central node selection and the clustering procedure. However, the original CNS does not consider the influence among the nodes and the importance of the division of the edges in networks. In this paper, an OCD algorithm based on a central edge selection (CES) algorithm for detection of overlapping communities of protein–protein interaction (PPI) networks is proposed. Different from the traditional CNS algorithms for OCD, the proposed algorithm uses community magnetic interference (CMI) to obtain more reasonable central edges in the process of CES, and employs a new distance between the non-central edge and the set of the central edges to divide the non-central edge into the correct cluster during the clustering procedure. In addition, the proposed CES improves the strategy of overlapping nodes pruning (ONP) to make the division more precisely. The experimental results on three benchmark networks and three biological PPI networks of Mus. musculus, Escherichia coli, and Cerevisiae show that the CES algorithm performs well.

The talent planning model and empirical research to the key disciplines in science and technology

With to the impact of economic globalization, the talent construction of key disciplines in science and technology should be administrated with humanism. An analysis of existing articles shows that the research of talent development mainly relates to the following aspects: cultivating objectives, cultivator, cultivation way, and evaluation criteria. In recent years, with the continuous improvement of education system in China and the increased awareness of talents, the talent construction of key discipline in science and technology has been greatly improved. With the actuality and circumstance analysis of the talent construction of key disciplines, a talent planning model is proposed to the key disciplines in science and technology. The proposed model is III-level tree structure, of which there are 2 I-level indexes, 8 II-level indexes and 23 III-level indexes. The Analytic Hierarchy Process is employed to determine the weights of talent planning indexes. This research will make the more scientific, systematic, strategic talent planning, and adapt to the development needs of key disciplines.

Partitioning Clustering Based on Support Vector Ranking

Support Vector Clustering (SVC) has become a significant boundary-based clustering algorithm. In this paper we propose a novel SVC algorithm named “Partitioning Clustering Based on Support Vector Ranking (PC-SVR)”, which is aimed at improving the traditional SVC, which suffers the drawback of high computational cost during the process of cluster partition. PC-SVR is divided into two parts. For the first part, we sort the support vectors (SVs) based on their geometrical properties in the feature space. Based on this, the second part is to partition the samples by utilizing the clustering algorithm of similarity segmentation based point sorting (CASS-PS) and thus produce the clustering. Theoretically, PC-SVR inherits the advantages of both SVC and CASS-PS while avoids the downsides of these two algorithms at the same time. According to the experimental results, PC-SVR demonstrates good performance in clustering, and it outperforms several existing approaches in terms of Rand index, adjust Rand index, and accuracy index.

Gene-Disease Interaction Retrieval from Multiple Sources: A Network Based Method

The number of gene-related databases has been growing largely along with the research on genes of bioinformatics. Those databases are filled with various gene functions, pathways, interactions, and so forth, while much biomedical knowledge about human diseases is stored as text in all kinds of literatures. Researchers have developed many methods to extract structured biomedical knowledge. Some study and improve text mining algorithms to achieve efficiency in order to cover as many data sources as possible, while some build open source database to accept individual submissions in order to achieve accuracy. This paper combines both efforts and biomedical ontologies to build an interaction network of multiple biomedical ontologies, which guarantees its robustness as well as its wide coverage of biomedical publications. Upon the network, we accomplish an algorithm which discovers paths between concept pairs and shows potential relations.