Journal of Information Systems and Telecommunication
Year:2017 | Volume:5 | Issue:2|Papers Count:8

Block modeling as a social structure discovery process needs to find and adopt a partitioning of actors to equivalent classes or positions. The best partitioning, naturally, must provide the closest estimation of network ties and show the most agreement with original network data. This interpretation of the best, leads to the structure with the most fitness to original network data. Finding this best partition vector can be formulated as an optimization problem and can be solved by Meta heuristic algorithms. In this paper, we use cuckoo search and genetic algorithm as a basis for comparison with cuckoo search. In addition to simple cuckoo search, we apply a hybrid cuckoo search algorithm to find the solution. The results of experiments through multiple samples reveals that while genetic algorithm shows the better performance in terms of convergence time and small iteration, the hybrid cuckoo search finds the better solutions than genetic algorithm in large iteration in terms of quality of solutions measured by fitness function. Furthermore, the hybrid cuckoo search shows no significant superiority over the simple cuckoo search, unless the large iteration numbered is used. In addition to block model problem, the proposed hybrid cuckoo search shows clear superiority over the greedy discrete PSO for community detection problem.

Coreference resolution is the problem of clustering mentions in a text that refer to the same entities, and is a crucial and difficult step in every natural language processing task. Despite the efforts that have been made to solve this problem during the past, its performance still does not meet today’s application requirements. Given the importance of the verbs in sentences, in this work, we tried to incorporate three types of their information on coreference resolution problem, namely, selectional restriction of verbs on their arguments, semantic relation between verb pairs, and the truth that arguments of a verb cannot be coreferent of each other. As a needed resource for supporting our model, we generate a repository of semantic relations between verb pairs automatically using Distributional Memory (DM), a state-of-the-art framework for distributional semantics. This resource consists of pairs of verbs associated with their probable arguments, their role mapping, and significance scores based on our measures. Our proposed model for coreference resolution encodes verb’s knowledge with Markov logic network rules on top of the deterministic Stanford coreference resolution system. Experiment results show that this semantic layer can improve the recall of the Stanford system while preserves its precision and improves it slightly.

Counting mitotic figures present in tissue samples from a patient with cancer, plays a crucial role in assessing the patient’s survival chances. In clinical practice, mitotic cells are counted manually by pathologists in order to grade the proliferative activity of breast tumors. However, detecting mitoses under a microscope is a labourious, time-consuming task which can benefit from computer aided diagnosis. In this research we aim to detect mitotic cells present in breast cancer tissue, using only texture and pattern features. To classify cells into mitotic and non-mitotic classes, we use an AdaBoost classifier, an ensemble learning method which uses other (weak) classifiers to construct a strong classifier.11 different classifiers were used separately as base learners, and their classification performance was recorded. The proposed ensemble classifier is tested on the standard MITOS-ATYPIA-14 dataset, where a 64×64 pixel window around each cells center was extracted to be used as training data. It was observed that an AdaBoost that used Logistic Regression as its base learner achieved a F1 Score of 0.85 using only texture features as input which shows a significant performance improvement over status quo. It is also observed that "Decision Trees" provides the best recall among base classifiers and "Random Forest" has the best Precision.

Natural disasters are an inevitable part of the world that we inhabit. Human casualties and financial losses are concomitants of these natural disasters. However, by an efficient crisis management program, we can minimize their physical and social damages. The real challenge in crisis management is the inability to timely receive the information from the stricken areas. Technology has come to the aid of crisis management programs to help find an answer to the problem. One of these technologies is wireless sensor network. With recent advances in this field, sensor nodes can independently respond to the queries from the users. This has transformed the processing of the queries into one of the most useful chapters in sensor networks. Without requiring any infrastructure, the sensor network can easily be deployed in the stricken area. And with the help of spatial query processing, it can easily provide managers with the latest information. The main problem, however, is the irregular shape of the area. Since these areas require many points to present them, the transmission of the coordinates by sensor nodes necessitates an increase in the number of data packet transmissions in the sensor network. The high number of packets considerably increases energy consumption. In related previous works, to solve this problem, line simplification algorithm s, such as Ramer-Douglas-Peucker (RDP), were used. These algorithms could lessen energy consumption by reducing the number of points in the shape of the area. In this article, we present a new algorithm to simplify packet shapes which can reduce more points with more accuracy. This results in decreasing the number of transmitted packets in the network, the concomitant reduction of energy consumption, and, finally, increasing network lifetime. Our proposed method was implemented in different scenarios and could on average reduce network’s energy consumption by 72.3%, while it caused only 4.5% carelessness which, when compared to previous methods, showed a far better performance.

This paper presented a multiple Distributed Learning Automata (DLA) random walk model for node classification on a social network task. The purpose of this work is to improve the accuracy of node classification in social network by using of DLA. When dealing with large graphs, such as those that arise within the context of online social networks, a subset of nodes may be labeled. These labels can indicate demographic values, interest, beliefs or other characteristics of the nodes. A core problem is to use this information to extend the labeling so that all nodes are assigned a label Due to the high accuracy of local similarity measures, in the proposed algorithms, we will use them to build the transition matrix. As a standard in social network analysis, we also consider these networks as graphs in which the nodes are connected by edges and the transition matrix is used as weight value of edges. Now we partition this graph according to labeled nodes. Every sub-graph contains one labeled node along with the rest of unlabeled nodes. Then corresponding DLA on each partition. In each sub-graph we find the maximal spanning tree by using of DLA. Finally, we assign label by looking at rewards of learning automata. We have tested this algorithm on three real social network data sets. The result of Experiments show that the expected accuracy of a presented algorithm is achieved.

Arithmetic units are essential in digital circuit construction, and the enhancement of their operation would optimize the whole digital system. Among them, multipliers are the most important operational units and are used in a wide range of digital systems such as telecommunication signal processing, embedded systems, and mobile technology. The main drawback to a multiplication unit is its high computational load, which leads to considerable power consumption and an increased area of silicon. This also reduces the speed, which negatively affects the digital host functionality. Estimating arithmetic is a new branch of computer arithmetic implemented by discarding or manipulating a portion of arithmetic circuits and/or intermediate computations. Applying estimated arithmetic in arithmetic units would improve the speed, power consumption, and the implementation area by sacrificing a slight amount of result accuracy. This article develops and analyzes an estimated truncated floating-point multiplier for single precision operands that is capable of compensating for errors to a desired level by applying the least significant columns of the partial product matrix. These errors are caused by removing a number of carry digits in the partial product matrix which make a direct contribution to rounding the floating-point numbers. The evaluation results indicate that the proposed method improves speed, accuracy, and silicon area, in comparison with common truncated multiplication methods.

Normally, the-state-of-the-art methods in field of object retrieval for large databases are achieved by training process. We propose a novel large-scale generic object retrieval which only uses a single query image and training-free. Current object retrieval methods require a part of image database for training to construct the classifier. This training can be supervised or unsupervised and semi-supervised. In the proposed method, the query image can be a typical real image of the object. The object is constructed based on Speeded Up Robust Features (SURF) points acquired from the image. Information of relative positions, scale and orientation between SURF points are calculated and constructed into the object model. Dynamic programming is used to try all possible combinations of SURF points for query and datasets images. The ability to match partial affine transformed object images comes from the robustness of SURF points and the flexibility of the model. Occlusion is handled by specifying the probability of a missing SURF point in the model. Experimental results show that this matching technique is robust under partial occlusion and rotation. The properties and performance of the proposed method are demonstrated on the large databases. The average of retrieval rate by the proposed method applied on Oxford landmarks and Corel dataset are 69.68% and 65.79%, respectively. Also, the average of ANMRR measure by the proposed method applied on Oxford landmarks is 0.223 and this criterion for Corel dataset is 0.269. The obtained results illustrate that the proposed method improves the efficiency, speeds up recovery and reduces the storage space.

In cognitive radio networks (CRN), resources available for use are usually very limited. This is generally because of the tight constraints by which the CRN operate. Of all the constraints, the most critical one is the level of permissible interference to the primary users (PUs). Attempts to mitigate the limiting effects of this constraint, thus achieving higher productivity is a current research focus and in this work, cooperative diversity is investigated as a promising solution for this problem. Cooperative diversity has the capability to achieve diversity gain for wireless networks. Therefore, the possibility of and mechanism for achieving greater utility for the CRN are studied when cooperative diversity is incorporated. To accomplish this, a resource allocation (RA) model is developed and analyzed for the heterogeneous, cooperative CRN. In the model, during cooperation, a best relay is selected to assist the secondary users (SUs) that have poor channel conditions. Overall, the cooperation makes it feasible for virtually all the SUs to improve their transmission rates while still causing minimal harm to the PUs. The results show a marked improvement in the RA performance of the CRN when cooperation is employed in contrast to when the CRN operates only by direct communication.