Dwijen Rudrapal

Voice Recognition and Authentication as a Proficient Biometric Tool and its Application in Online Exam for P.H People

Voice Recognition is a fascinating field spanning several areas of computer science and mathematics. Reliable speech recognition is a hard problem, requiring a combination of many complex techniques; however modern methods have been able to achieve an impressive degree of accuracy. On the other hand, today, most of the companies or institutes are conducting their examinations online to be a part of this best ever growing world. In this system user can give any available examination at any accessible center as per his/her choice and authority also can condense manpower and process delay overhead. This paper offers one way to conduct online examination for physically challenged people who can use their voice only to register and attend the examination. In addition, in the course of this paper it has been tried to authenticate one registered user and to make the authentication process persistent throughout the examination interlude.

Analysis and Evaluation of Keystroke Duration of User’s Typing as a Distinctive Measure of Recognition

This paper illustrated one unique measure of keystroke patterns of user typing to prevent insider threats with combination of existing keystroke credentials. While insider threats may never be eliminated, it is worth noting that risk-mitigating counter measures exist. Every organization’s security architecture must take insider threat into account. As reiterated trough out the security industry, defense in depth is the essential. Defense in depth requires security professional to take a layered approach to protecting information. In essence, we should never rely on just one security method or technology. Only user ID and passwords are not enough strong now a day to protect workstations from unauthorized access. Being remarkable advantages and to support multilevel authentication biometrics may be strong add-on. Recent research on behavioral biometrics i.e., keystroke dynamics proves its efficiency for stronger authentication. With numerous advantages keystroke pattern credential are employing in many organization to establish layered security mechanism. Work of last decade on keystroke dynamics also reveals some drawbacks. So research is going on to find more traits of user typing for more efficient authentication. Through this paper we proposed and experimentally proved one measure of keystroke pattern which is distinctive for each user and can make keystroke pattern credential stronger.

Improvisation of Biometrics Authentication and Identification through Keystrokes Pattern Analysis

In this paper we presented one fresh approach where the authentic users typing credentials are combined with the password to make authentication convincingly more secure than the usual password used in both offline and online transactions. With the help of empirical data and prototype implementation of the approach, we justified that our approach is ease of use, improved in security and performance. In normal approach different keystroke event timing is used for user profile creation. Keystroke latency and duration is inadequate for user authentication, which motivates exploring other matrices. In this paper we proposed combination of different matrices and calculation of degree of disorder on keystroke latency as well as duration to generate user profile. Statistical analysis on these matrices evaluates enhanced authentication process.

Performance Analysis of Hierarchical and Flat Network Routing Protocols in Wireless Sensor Network Using Ns-2

networks have the characteristics of fault tolerance and rapid development make them very promising in the above mentioned field. The main objective of this paper is to analyze the performance of two categories routing protocols i.e.

Measuring the Limit of Semantic Divergence for English Tweets.

In human language, an expression could be conveyed in many ways by different people. Even that the same person may express same sentence quite differently when addressing different audiences, using different modalities, or using different syntactic variations or may use different set of vocabulary. The possibility of such endless surface form of text while the meaning of the text remains almost same, poses many challenges for Natural Language Processing (NLP) systems like question-answering system, machine translation system and text summarization. This research paper is an endeavor to understand the characteristic of such endless semantic divergence. In this research work we develop a corpus of 1525 semantic divergent sentences for 200 English tweets.

Quotology - Reading Between the Lines of Quotations

A quote is short eloquent sentence/s drawn from long experience. Quotes are full of poetry, abstraction in the form of restrained information, hidden significance and objective, and pragmatic twists. While a quote explanation always in elaborate form with more sentences than the quote, both of them convey the same concept or meaning. However, systematic study to understand linguistic structures and interpretation of quotes has not received much research attention till date. To this end we have developed a corpus of English quotes and their interpretations in elaborated forms. Finally, we proposed an automatic approach to recognize Textual Entailment (TE) between English quote and its explanation where quote has been considered as the text (T) and its explanation/interpretation has been considered as hypothesis (H). We have tested various linguistic features including lexical, syntactic, and semantic cues and also tried word-to-vector similarity measure using deep learning approach on quote-explanation to identify TE relation.

OCR for devanagari numerals using zonal histogram of angle

Abstract Optical Character Recognition (OCR) system is used to generate the textual representation of handwritten or printed text. Many research works are going on in the field of OCR over the past few decades for most of the Indian scripts. Devanagari one of the most spoken languages in the world as well as India. Lack of a robust OCR system for Devanagari script is still there even after so much research. The aim of this paper is to make an OCR that could classify handwritten Devanagari numerals. This paper proposes an OCR based on Histogram of the angle made by a dark pixel with the zonal center of mass. This feature bags the angle made by each dark pixel in a zone about its center of mass. This newly extracted feature was used to train various classification algorithms like K-Nearest Neighbor, SVM, Linear SVM, Random Forest, Decision Tree, Gradient Boosting, Gaussian Naive Bayes. We reported an efficiency of each algorithm based on the new feature. Our experiment result shows that the Random Forest Model outperforms over the other algorithms and reports an efficiency of 92.57%.

Analysis of Color Moment as a Low Level Feature in Improvement of Content Based Image Retrieval

In the recent past, the rapid intensification of the Internet has significantly enhanced the quantity of image collections accessible owing to the simplicity with which images are being formed or stored. CBIR (Content Based Image Retrieval) phenomenon is therefore highly encouraged by this requirement of unbeaten and proficient exploration of the large image databases. Consequently, the low level feature extraction of the visual contents of an image and their analysis is very significant in terms of CBIR. These low level features can be colour, texture and shape features. As Colour based image retrieval procedure is the trendiest of all these feature extraction algorithms, hence in this paper the color moments of the Hue, Saturation, and Value (HSV) component images in HSV color space are used as feature extraction algorithm. After the successful calculation of features for extraction, similarity computation is done using Euclidean Distance in between the test image and object images and finally the image retrieval is done. Analysis of this paper shows that the training time required for individual image, as well as, all the images in the database is very small which provides instantaneous retrieval. The estimation of the proposed approach is conceded out using the standard precision, recall and f-score measures, and the experimental results demonstrate that the proposed method has higher accuracy and retrieval rate than the conventional methods.

A system to filter unsolicited texts from social learning networks

In the present day scenario online social networks (OSN) are very trendy and one of the most interactive medium to share, communicate and exchange numerous types of information like text, image, audio, video etc. All these publicly shared information are explicitly viewed by connected people in the blog or networks and having an enormous social impact in human mind. Posting or commenting on particular public/private areas called wall, may include superfluous messages or sensitive data. Information filtering can therefore have a solid influence in online social networks and it can be used to give users the facility to organize the messages written on public areas by filtering out unwanted wordings. In this paper, we have proposed a system which may allow OSN users to have a direct control on posting or commenting on their walls with the help of information filtering. This is achieved through text pattern matching system, that allows users to filter their open space and a privilege to add new words treated as unwanted. For experimental analysis a test social learning website is designed and some unwanted words/texts are kept as blacklisted vocabulary. To provide control to the user, pattern matching of texts are done with the blacklisted vocabulary. If it passes then only text can be posted on someones wall, otherwise text will be blurred or encoded with special symbols. Analysis of experimental results shows high accuracy of the proposed system.