By adopting the concepts of cryptography and image processing we will provide security for data through the image. This paper proposes a cryptographic technique by considering functionality for the security of data and image. By using triple DES algorithm data encryption and decryption is done .Later through LSB technique encryption of data, the cipher data can be stored into the image. After storing data the image can be encrypted by using DIT (Data Inverse technique).Once the encryption of image is completed the image compression is done using Arithmetic compression and decompression technique. The proposed architecture provides data confidentiality and improves the performance. These days transferring data and image through the network without losing data integrity are a complex task to provide security of data and image, we can use the cryptography techniques, image encryption, and decryption and compression techniques.

Keywords: LSB, DIT, ETC, Arithmetic Compression and decompression technique, image encryption and decryption.

The paper proposes a new initiate to enhance the existing online multi-modal distance metric learning (OMDML) with a new feature of extension to solve the Image ambiguity issue using Conditional Random Field (CRF) Algorithm. The main intention of proposing this model of system is to annotate/tag the images with some manually defined concepts for learning an inherent space, using visual and contextual features. More particularly, by making the system to feed the latent vectors into existing classification representations, it can be enforce for application of image annotation, which is considered as the needed issue in image retrieval. As an extension to the available model, we recommend and add the entity feature of the problem of solving the ambiguity. The Conditional Random Filed Algorithm model is used for training the system and results of reinforced online multi-modal distance metric learning system provides an better outcome of content based image retrieval model. This solution is the future enhancement where the contribution of providing more accuracy to the proposed system by enhancing using ambiguity resolving problem.

The development of a miniaturized two-element printed MIMO antenna with a triple-band performance for WiFi and WiMAX operability is presented in this paper. The proposed antenna configuration incorporates two small-sized square copper patches printed on a standard FR-4 substrate with an overall size of 40×27.2 mm2. The two patches are furnished with slots into the form of split-ring resonators and fed from two extended microstrip lines with trimmed ground planes to allow for triple-band operation. A neutralizing line connecting the two patches is also incorporated into the antenna structure to improve isolation. Double-sided metamaterial grid is included to fill the space between the two antenna elements providing thereby for enhanced radiation efficiency and improved isolation for the second band. Practical results clearly depict the triple band performance with a slight slip in the second and third bands form the ones predicted by simulation covering the following ranges of frequencies: (2.4-2.485) GHz, (5.1-5.35) GHz for Wi-Fi (IEEE 802.11) and (3.35-3.8) GHz for Wi-max (IEEE 802.16.2005).
Keywords— Printed Antenna, Metamaterial, Neutralizing Line, Split Ring Resonator, MIMO, WiFi, WiMAX

A recommender system is a system performing information filtering to fetch information items such as movies, music, books, news, images, web pages and tools for a user. This information is filtered so that it is likely to be used by the user. The aim of a recommender system is often to "help consumers learn about new products and desirable ones among myriad of choices" this would enable in improving the recommendation quality, accuracy and diversity. In this paper a item popularity based ranking algorithm is utilized. By using this algorithm the best selling items are recommended to each user. In this paper different types of ranking techniques have been considered. The techniques used are Item Popularity-Based Ranking, Reverse Predicted Rating Value, Item Average Rating, Item Absolute Likeability and Item Relative Likeability. These techniques can generate recommendations that have substantially and higher aggregate diversity across all users while maintaining optimal levels of recommended accuracy.

Keywords— Recommender systems, Recommendation accuracy, Recommendation diversity, Ranking functions, collaborative filtering.

Human Computer interaction is one of the basic requirements of people in our time, humanity has been interacting with computers in every possible way, this work focuses on communication capabilities using brain signals. Such technology offers a great potential especially for people suffering from limited communication capabilities and require assistive devices. This type of communication schemes is usually referred to as brain computer interface (BCI), which is a system that establishes a direct link with the human brain without any muscular movement by translating neurons electrical activity into useful commands. However, such system usually suffers from low information transfer rate as well as low accuracies. This represents a challenge when used to deliver information of large set of characters as in natural language. This paper aims at designing a spelling application based on brain computer interface using Arabic Letters. The BCI adopted in this work uses event-related potential (ERP) by stimulating the neurons into generating detectable signals. When the subject pays attention to a stimulus presented in real-time signal is recoded, processed and classified into commands performed by the spelling application. The system has been tested under various parameters and conditions, the proposed method has indeed proved to be operational and adequate.

Keywords— HCI, BCI, P300, Arabic Speller.

The women's in the age of reproduction, the heterogeneous, complex and multifaceted disease called as PCOS (Polycystic ovary syndrome) affects about 5 to 10 % of the women. Metabolic syndrome (frequent metabolic traits) called as hypertension, hyperinsulinemia, abdominal obesity and dyslipidemia and these are characterized along with the resistance of insulin like chronic anovulation, hyperandrogenism and polycystic ovaries which ends with the serious diseases like coronary disease, endometrial hyperplasia and type 2 diabetes mellitus these are the long term consequences. For decision making purpose in PCOS phenotype (i.e to discover the main cause of PCOS with many number of attributes (tests)), an approach that used to store and manipulating the valued data for predicting the causes earlier and this process in information technology is called as data mining. In the previous research paper, a novel framework is proposed for the prediction of PCOS disease among women's. In this paper, a hypothesis test is made on the proposed methodologies to compare the effectiveness of the proposed techniques with existing techniques.

Keywords— PCOS, Hypothesis, Chi-Square test.