"National Conference on Recent Advances in Computer Sciences NCRACS 2014



IJSER Research Group https://www.ijser.org/forum/index.php Register for IJSER Research Forum      IJSER Xplore IJSER Xplore Research Paper Database
 

Implementation of an Adaptive Antenna Array Algorithm for Anti-Jamming Techniques[ ]


A smart antenna has the capability of suppressing jamming signal, they can improve Signal to Interference plus Noise Ratio (SINR). Antennae array processing utilizes information regarding locations of signal to aid in interference suppression and signal enhancement and is considered promising technology for anti-jamming. Simulation results are presented to compare the ability of Least Mean Square (LMS), Recursive Least Squares (RLS) and the conjugate gradient method algorithms to form beam in the direction of desired signal and place null to cancel out interference signal with the assumption of angle of desired signal and angle of interference. Analysis of these algorithms on SNR and number of iterations needed to get desired signal that is minimize the error faster than other adaptive algorithms are analyzed.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
FORGERY IMAGE DETECTION BASED ON ILLUMINATION COLOR CLASSIFICATION WITH ADVANCED SKIN COLOR AND EDGES[ ]


In recent days, photographs have been used as evidence in courts. Photographers are able to create composites of analog pictures, this process is very time consuming and requires expert knowledge. Today, Powerful digital image editing software makes image modi?cations straightforward. This undermines our trust in photographs. In this paper, one of the most common forms of photographic manipulation, known as image composition or splicing is analysed .A forgery detection method that exploits subtle inconsistencies in the color of the illumination of images. The proposed approach is machine-learning based and requires minimal user interaction. The technique is applicable to images containing two or more people and requires no expert interaction for the tampering decision. Here, the existing work can be extended by using advanced face detection method using skin tone information and edges . A lighting insensitive face detection method based upon the edge and skin tone information of the input color image is proposed. From these illuminant estimates, we extract texture- and edge-based features which are then provided to a machine-learning approach for automatic decision-making.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Ontology Based Approach For Instance Matching[ ]


One of the important barrier that hinders achieving semantic interoperability is ontology matching. Instance-based ontology matching (IBOM) or concept based ontology matching(CBOM) uses the extension of concepts, the instances directly associated with a concept, to determine whether a pair of concepts is related or not. Practically, instances are often associated with concepts of a single ontology only, rendering IBOM rarely applicable. This is achieved by enriching instances of each dataset with the conceptual annotations of the most similar instances from the other dataset, creating artificially dually annotated instances. We call this technique concept based ontology matching by concept enrichment (CBOMbCE). We are using the instance matching process with web crawlers mediating three world’s leading publishers such as Oxford, ScienceDirect and Springer. We are obtaining keywords from the articles of these four journals which acts as the instances. We are collecting all possible journals available in these three websites since the access permission of these three journals can be restricted to some constraints within it. After searching and finding keywords those instances are matched with their ontology creation and further enrichment of instances. Through this technique we will obtain instances that are uncommon among two datasets.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
An Efficient Intrusion Detection System for EAACK Scheme in MANETs using DSAB[ ]


The movement from wired network to wireless network has been a trend in the past few years. Comparing to other wireless networks MANETs is important one.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Improved Project Gradient Method for Glioma Diagnosis[ ]


Medical Image Processing is one of the most challenging and emerging topics in today’s research field. Processing of Magnetic Reson-ance Spectroscopic Imaging (MRSI) is one of the parts in this field. In recent years, multispectral MRI has emerged as an alternative to Ultra-sound (US) image modality for clear identification of tumors. In order to analyze a disease, Physicians consider MR imaging modality is the most efficient one for identification of tumors present in Brain. Therefore, analysis on MR imaging is required for efficient disease diagnosis. The nosologic images of the brain using magnetic resonance spectroscopic imaging (MRSI) data in an unsupervised way is created to differentiate various tissue patterns of glioma (Brain Tumor). Different tissue patterns are identified from the MRSI data using improved project gradient method for nonnegative matrix factorization and are then coded as different primary colors (i.e. red, green, and blue) in an RGB image, so that mixed tissue regions are automatically visualized as mixtures of primary colors. Nosological images is useful in assisting glioma diagnosis, where several tissue patterns such as normal, tumor, and necrotic tissue can be present in the same voxel/spectrum. Error-maps based on linear least squares estimation are computed for each nosologic image to provide additional reliability information, which may help clinicians in decision making. Thus detection and extraction of brain tissue from MRI image is effectively done for glioma diagnosis

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Routing Overhead Reduction in MANETs[ ]


Routing is one of the challenging issues in Mobile Ad hoc NETw orks (MANETs). Broadcasting is the fundamental and eff icient data dissemination mechanism for route discovery in reactive routing protocols of Mobile Ad hoc NETw ork (MANET). This causes the problem called the broadcast storm problem w hich results in redundant retransmission and adds to routing overhead. There are many approaches proposed to solve the problem; but none of them addresses the problem effectively. This paper proposes a new mechanism that has probabilistic rebroadcast based on neighbor coverage for the routing overhead reduction. This proposed mechanism w ill reduces the packet retransmission and thus reduce the routing overhead. This approach combines the advantages of probabilistic mechanism and neighbor area coverage based approach. This new mechanism can improve the performance of broadcasting in various network scenarios. This approach is simple and can be implemented in NS-2.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Removal Of Blurred And Illuminated Face Image With Different Poses[ ]


Face recognition has been researched field of computer vision for the past twenty years. This work have addressed the matter of recognizing blurred and poorly well-lighted faces. The set of all pictures obtained by blurring a given image may be a umbellate set given by the umbellate hull of shifted versions of the image. Supported this set-theoretic characterization, the work planned a blur-robust face recognition rule DRBF. This rule will simply incorporate previous information on the kind of blur as constraints. Determining the low-dimensional linear topological space model for illumination, the work showed that the set of all pictures obtained from a given image by blurring and ever-changing its illumination conditions may be a bi-convex set. Again, supported this set-theoretic characterization, this work planned a blur and illumination strong rule IRBF. The face below completely different create are often detected and normalized by mistreatment transformation parameters to align the input create image to frontal read. When finishing the said create social control method, the ensuing final image undergoes illumination social control. This is often performed mistreatment the SQI rule. Then face are often recognized mistreatment incorporating blur and illumination by classifying coaching and testing knowledge by mistreatment SVM.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Image Super Resolution Using Multiple Kernel Learning[ ]


Learning-based approaches for image super-resolution (SR) have attracted the attention of researchers in the past few years.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Searching and Classification of List of Keywords and URLs Using SVM Classifier[ ]


Internet forums are in the main used for discussions where users can request and exchange data with others. Forum contains form of top-ics associated with user interest. The goal of net travel is to retrieve forum content with smallest overhead. The target of the net crawlers is to send users from entry page to thread page. The thread page contains data content relevant to user question. The projected net travel primarily based on the mechanism that supports multiple keyword based retrieval using ontology. Ontology helps the crawler to extract and aggregate information from a specific domain. This proposed methodology addresses the uniform resource locator kind recognition drawback to get rid of the duplicate and unwanted pages. The strategy is evaluated exploiting SVM classifier and located to be higher than existing systems.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
UP-GROWTH++ ALGORITHM FOR MINING HIGH UTILITY ITEMSETS[ ]


The efficient discovery of high utility itemset from large transaction database is a crucial task of data mining. In past, many relevant algorithms have been presented. These algorithms surface the problem of generating large number of candidate itemset and thus degrade the mining performance in terms of execution time and space. In this paper, three algorithms are presented, such as EUP-Growth (utility pattern) for mining high utilty itemset for pruning candidate itemsets. In these algorithms, compact tree structure (EUP-Tree) is used for discovering the useful itemset so that candidate item is generated with only two scan of database. The performance of EUP-Growth and is compared with the state of-the algorithm for both real and synthetic data sets. Experimental results shown that the proposed algorithm reduce the number of candidates effectively but also outperform other algorithms substantially in terms of runtime even when databases contain lots of long transactions

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Mitigating Snoop-Forge-Replay Attack by Integrating Text-Based and Language-Based Traits with the Keystroke Verification System[ ]


A new attack called the snoop-forge-replay attack is presented on keystroke-based continuous verification systems. The snoop-forge-replay is a sample-level forgery attack and is not specific to any particular keystroke-based continuous verification method or sysem. It can be launched with easily available keyloggers and APIs for keystroke synthesis. Our results ffrom 2460 experiments show that: 1)the snoop-forge-replay attacks achieve alarmingly high error rates compared to zero-effort imposter attacks, which have been the de facto standard for evaluating keystroke-based continuous verification systems; 2)four state-of-the –art verification methods, three types of keystroke latencies, and 11 matching-pair settings(a key parameter in continuous verification with keystrokes) that is examined here were suspectible to the attack; 3)the attack is effective even when as low as 20 to 100 keystrokes were snooped to create forgeries.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Real Time Sentiment Classification Using Unsupervised Reviews[ ]


Sentiment classif ication is an important task in everyday life. Users express their opinion about their product, movies and so on. All the web page contains reviews that are given by users expressing different polarity i.e. positive or negative. It is useful for both the producer and consumer to know what people think about the particular product or services based on their reviews. Automatic document classif ication [2],[3] is the task of classifying the reviews based on the sentiment expressed by the reviews. Sentiment is expressed differently in different domains. The data trained on one domain cannot be applied to the data trained on another domain [6]. The cross domain sentiment classif ication overcomes these problems by creating thesaurus for labeled data on the target domain and unlabeled data f rom source and target domains. Sentiment sensitivity is achieved by creating thesaurus. The sentiments cannot be analyzed for sentence and the data to be trained on a particular domain. The proposed method focus on unsupervised method of using Tw itter, the most popular micro blogging platform, for the task of Opinion analysis. A sentiment classif ier, that is able to determine positive, negative and neutral sentiments for a tw itter website reviews.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------