Modern organizations keep up with the fast changing technology through successful information systems development and integration to automate their business operations. Among others, this can be achieved through effective requirements elicitation which a sub-phase of requirements engineering. It is recognized as one of the most critical, knowledge-intensive phase therefore determining the overall success or failure of the information system. In a setup whereby users are heterogeneous by nature, requirements elicitation becomes quite challenging. To improve the effectiveness of the requirements elicitation process from this type of users, the analysts should determine the factors that affect this process which are unique to this type of users. To address this issue, we have conducted research in two Kenyan public universities whereby information systems users are heterogeneous in nature. The objective of the study was to determine the factors that affect requirements elicitation process among heterogeneous user groups of information systems. Nine factors were identified by using multiple regression analysis  namely, users’ availability, users’ and analysts’ diverse cultural background, users’ geographical distribution, communication issues, users’ awareness and training on the process, technique used, institutional politics and bureaucracy, requirements elicitation policy and user's cooperation & motivation. These factors form a practical guide that can help analysts to consider before carrying out requirements elicitation from heterogeneous users.

Keywords: Requirements elicitation, requirements elicitation models, requirements engineering, software development, heterogeneous users

Routing protocols are the heart and soul of internetworking, routers uses routing protocols as their algorithms to decide the best path from a given source to a destination. That being said it is very critical to choose the right routing protocol for the topology as many parameters exists such as throughput, delay and QoS. Network convergence on the other hand is the state of set of routers having the same topological awareness about the network.
In this paper I will discuss the network convergence activity and time if the given routing protocols (EIGRP, OSPF, and RIP) are being simulated. Also I will discuss the convergence time and how implementing each protocol affects our simulation.

Keywords-component; routing, convergence, EIGRP, OSPF

 

This paper discusses a statistical algorithm to detect DOS attacks on computer net- works. DOS attacks hamper the network by making resources unavailable to genuine users. The algorithm presented here use SNMP data in order to detect incoming flooding attack on a computer or network. The data to be monitored depends on the class of flooding attacks that is intended to be detected. In this paper we focus on detecting ICMP, UDP, TCP and IP type of flooding attacks on various network interfaces.

 

-In the era of technology, Editing, manipulation, copying of digital multimedia has become prevalent and becoming a serious issue in the present scenario. Digital watermarking is one of the solution to deal with this problem. It is a technique in which pattern of bits are embedded into the digital data in such a way that it do not hinder the actual data. Thus it maintains its reliability. Digital watermarking can be applied in the variety of application areas such as copyright protection, fingerprinting, broadcasting monitoring, copy protection, medical application. In the present review paper, the author is pointing out challenges and issues which occurred while this technique is applied.

Keywords: Digital Watermark, DWT, Imperceptibility, Robustness

 

MANET consists of mobile platforms have a router with multiple hosts and wireless communications devices. Energy savings and security is the major problem in all kind of the networks. MANET is a network, which is very popular due to its unique characteristics from all the other types of networks. MANET is a network having tiny light weighted nodes, with no clock synchronization mechanisms. Generally, in this type of network the exhaustion of energy will be more and as well, the security is missing due to its infrastructure less nature. Due to the lack of energy, the link failure may occur and the network lifetime also gets affected. Similarly, the node causes cheating during the transmission process in the network Most MANET routing protocols are vulnerable to attacks that can freeze the whole network. Thus these may affects the performance of the network. To overcome these problems, we propose a new secured energy aware routing algorithm called SRECRP. This algorithm holds two mechanisms. Initially it makes all the active state nodes to sleep when not in use and then finds the energy based efficient path for reliable data transmission. Secondly, provides the security against attacks using a secured mechanism. By simulation based studies, we show that this algorithm effectively provides how its achieves higher security, less energy utilization, less overhead and less end to end delay.

Keywords—SRECRP, Energy, Link failure, MANET, Network, Security.

 

In India , online matrimonial websites plays a major role being trusted by millions of Indians globally. At the same time these act as the playground for the fraudsters and imposters duping thousands of victims by posting false and fraudulent profiles every year. Analytics is the future of big data because only transforming data into information gives them value and can turn data for social benefit. This paper investigates the benefits of Big Data technology and main methods of analysis that can be applied to the particular case of fraud detection in Indian Matrimonial Websites.

Keywords—Big Data, Matrimonial Fraud Detection, Hadoop, Social Networks

 

Word spotting is a technique which can extract the text from input image. Here, we implemented on scanned Tamil land documents. Using Gabor feature, we extract the feature values for the input image. The main goal is recognize the text from the document using K nearest neighbor classifier. The features were calculated and the features were combined.  Using these features, we can classify and recognized a text using an unsupervised classifier. The Classifier needs training of a database. The database consists of some set of words (printed or scanned). For the classification of the feature vector the k nearest neighbor Classifiers were employed.

Keywords: word spotting, recognition, text, feature, k nearest neighbor.