Aboamama Atahar Ahmed

Content-based Information Retrieval Techniques Based on Grid Computing: A Review

Abstract Distributed information retrieval methods are growing rapidly because of the rising need to access and search distributed digital documents. However, the content-based information retrieval (CBIR) is concentrated to extract and retrieve the information from massive digital libraries, which require a huge amount of computing and storage resources. The grid computing provides the reliable infrastructure for effective and efficient retrieval on these large collections. In order to build an effective and efficient CBIR technique, varieties of architectures were developed based on grid technologies. The goal of such architecture is to solve interoperability and heterogeneous resource issues, and increase the efficiency and effectiveness of information retrieval (IR) techniques by harnessing the grid computing capabilities. This paper reviews and analyzes latest research carried out in the domain of large-scale dataset IR based on a grid. The evaluation is based on scalability, response time, scope, data type, search technique, middleware, and query type. The contribution is to illustrate the features, capabilities, and shortages of current solutions that can guide the researchers in this evolving area.

OPTIMIZED QUERY FORWARDING FOR RESOURCE DISCOVERY IN UNSTRUCTURED PEER-TO-PEER GRIDS

One of the challenges of resource discovery in unstructured peer-to-peer grid systems is minimizing network traffic. The network traffic arises by query messages that are broadcasted to other peers in order to find the appropriate resources. Blind search methods that are employed in such systems do not work well because every specific query generates high query traffic, which quickly overwhelms the network. Informed search methods usually use recorded history of previous queries to decide where the new queries should be sent. Such methods can reduce network traffic but do not consider the path length. In this study, a method was proposed in which both the path length and network traffic are considered. This approach reduces the hop numbers and prevents massive flooding of query messages. To do this, it selects optimum neighbor peer(s) in order to optimize query forwarding. The proposed approach uses statistical tables that are obtained from recorded history of previous queries. Then a genetic algorithm is applied to these statistical tables to find the optimum neighbor peer(s). The proposed approach showed that query forwarding through the optimum neighbor peer(s) has a greater probability of finding a requested resource with lower hop numbers. This method was compared with random walk and flooding approaches. It was observed that the network traffic remarkably decreased in comparison to a flooding approach, whereas it was similar to the results obtained by a random walk method. Moreover, this method provided a higher success rate in comparison to the random walk approach, whereas it was similar to the results obtained by a flooding method.

A New Model for Resource Discovery in Grid Environment

Resource discovery is the most important challenge of Grid systems. Lately many researchers tried to improve indexing and searching methods. To this aim, there are several factors affecting improvement such as time and message complexities, scalability, dynamicity, reliability, support of different types of queries, and so on.

C-HLRD: a category-based hierarchical localisation technique for resource discovery in grid environments

Grid resource discovery which is the mechanism of indexing and locating the resources affects significantly the efficiency of grid systems. Some specific characteristics of grid environments such as dynamicity and large-scale make the resource discovery a challenging task. Different approaches for grid resource discovery have been proposed by researchers in recent years. This paper proposes a resource discovery technique highly adapted to large-scale grid environments with the aim of decreasing the message load and response time. It uses a category-based hierarchical structure to index and locate the resources. By this way, the resource query is localised rapidly and efficiently. The indexing structure and processes are implemented in GridSim simulation environment. The results of experiments showed significant decrease in the response time and message load during the search process especially in large scales.

An Optimized Search Algorithm for Resource Discovery in Peer to Peer Grid

One of the challenges for resource discovery in unstructured peer to peer grid is the minimizing of network traffic that is produced by query messages which are broadcasted to other nodes to find appropriate resources in the grid. These methods do not work well because each specific query generates a large amount of network traffic, and network is quickly saturated by the query messages. This study is proposed a genetic algorithm to find the required resources in peer to peer grid. This method prevents the massive flooding of the network traffic and decreasing query messages by using the restricted and optimized flooding. We compare our method with traditional approaches. Obtained results show that proposed method decreases the network traffic.