International Journal of Mathematical Sciences and Computing @ijmsc
Статьи журнала - International Journal of Mathematical Sciences and Computing
Все статьи: 258
Cryptographic Security using Various Encryption and Decryption Method
Статья научная
Fast development in universal computing and the growth in radio/wireless and mobile strategies have led to the extended use of application space for Radio Frequency (RFID), wireless sensors, Internet of things (IoT). There are numerous applications that are safe and privacy sensitive. The increase of the new equipments has permitted intellectual methods of linking physical strategies and the computing worlds through numerous network interfaces. Consequently, it is compulsory to take note of the essential risks subsequent from these communications. In Wireless systems, RFID and sensor linkages are extremely organized in soldierly, profitable and locomotive submissions. With the extensive use of the wireless and mobile devices, safety has therefore become a major concern. As a consequence, need for extremely protected encryption and decryption primitives in such devices is very important than before.
Бесплатно
Cyber Bullying Detection and Classification using Multinomial Naïve Bayes and Fuzzy Logic
Статья научная
The advent of different social networking sites has enabled people to easily connect all over the world and share their interests. However, Social Networking Sites are providing opportunities for cyber bullying activities that poses significant threat to physical and mental health of the victims. Social media platforms like Facebook, Twitter, Instagram etc. are vulnerable to cyber bullying and incidents like these are very common now-a-days. A large number of victims may be saved from the impacts of cyber bullying if it can be detected and the criminals are identified. In this work, a machine learning based approach is proposed to detect cyber bullying activities from social network data. Multinomial Naïve Bayes classifier is used to classify the type of bullying. With training, the algorithm classifies cyber bullying as- Shaming, Sexual harassment and Racism. Experimental results show that the accuracy of the classifier for considered data set is 88.76%. Fuzzy rule sets are designed as well to specify the strength of different types of bullying.
Бесплатно
Data Privacy System Using Steganography and Cryptography
Статья научная
Data privacy is being breached occasionally whether in storage or in transmission. This is due to the spate of attack occasioned by the movement of data and information on an insecure internet. This study aimed to design a system that would be used by both sender and receiver of a secret message. The system used the combination of Steganography (MSB) and Cryptography (RSA) approaches to ensure data privacy protection. The system generates two keys: public and private keys, for the sender and receiver to encrypt and decrypt the message respectively. The steganography method used does not affect the size of cover image. The software was designed using python programming language in PyCharmIDE. The designed system enhanced the security and privacy of data. The results of this study reveal the effectiveness of combination of steganography and cryptography over the use of either cryptography or steganography and other existing systems.
Бесплатно
Data- and Workflow Customer-Oriented Software Process Models
Статья научная
This paper presents a dataflow model to control the flow of data in each phase of a customer-oriented software process model. In addition, we suggest a workflow model to describe the transaction between the model phases, and a role model to govern the personnel participation and roles. Our goal is to develop models that involve the customer frequently and effectively during project development. Testing the models using CHAOS Report and shows that our models are capable of achieving this goal.
Бесплатно
Deep Classifier for Conjunctivitis – A Three-Fold Binary Approach
Статья научная
Alterations in environmental and demographic equations have resulted in phenomenal rise of human centric diseases, ocular being one of them. Technological advancements have witnessed early diagnosis of much of the previously un-ciphered diseases. This paper addresses two research questions (RQs) with the study being focused on conjunctivitis (the most prevalent eye ailment in adults as well as minors). The motive of both the RQs rests in implementing three state-of-the art deep learning framework for classification of the ocular disease and validation of the frameworks. Validation of the frameworks is seconded by improvised proposals for enhancements. RQ1 establishes and validates whether the three off the shelf Deep Learning frameworks VGG19, ResNet50, and Inception V3 properly classify the disease or not. RQ2 analyses the effectiveness of each classifier with further enhancement proposals. The algorithms were implemented on 210 images and generated an accuracy of 87.3%, 93.6%, and 95.2% for VGG19, ResNet50, and Inception V3 using Adam optimizer, with slightly variant results when applying Adadelta optimizer. These results were typical of the classification frameworks with enhancements. With pervasive penetration of Artificial Intelligence in healthcare, this paper presents the efficacy of Deep Learning Frameworks in conjunctivitis classification.
Бесплатно
Статья научная
This paper outlines the application of various classification methods and analytical techniques to identify a potential fraud. The aim of this document is to showcase the usefulness of such classification and analytical techniques for fraud detection. Considering the fact that there are hundreds of statistical methods and procedures to perform such analysis. In this paper, I would like to present a hybrid fraud detection method by using the Bayesian Classification technique to identify the risk group; followed by Benford's Law (The Law of First Digit) to detect a fraudulent transaction done by the identified risk group. Though this analysis focuses on the healthcare dataset, however, this technique can be replicated in any industry setup. Also, by adding the Voice of the Customer data to these classification and statistical methods, makes this analysis even more powerful and robust with improved accuracy.
Бесплатно
