Asian Journal of Research in Computer Science

In the Technological National of Mexico in San Luis Potosí (TecNM of SLP), Studies in computer systems in Engineering are taught, with a distance education model. However, it is detected that sometimes some classes are lost, which implies little effectiveness of the model. In the present study, a collection of data on connection effectiveness was carried out, to propose a reinforcement in the courses taught. Teachers were surveyed between January 2018 and December 2018, for each session and the results showed that 96.42% of the sessions were connected and were of good quality. Even so, it was proposed the realization and use of virtual courses to which the students could have asynchronous and timeless access, as an extra support, in case the students do not attend the campuses, they want to return to the class to clarify any doubt or review the class for preparation of practical work or evaluation.

Web security is a critical aspect for many web-based applications, along its research track, keystroke dynamics techniques have attracted broad interests due to their high efficiency in security. In this paper, the aim was to come out with a keystroke login system that overcomes the typical challenges associated with keystroke dynamics and improves on password security but with focus on irritability nature of keystroke dynamics based systems. Specifically, we proposed two stages user matching method, training/enrolment phase of users and authenticating registered users with previously stored data. Furthermore, the proposed algorithm added dwell, flight times and multiplied by the locate time to get the upper and lower bounds. Moreover, the uniform differences between the bound timings were calculated to further enhance security. Experimental results show that the proposed keystroke dynamics approach used in augmenting password security emerged to be superior as compared to existing customary distance metrics.

Soft-computing techniques for fire safety parameter predictions in flammability studies are essential for describing a material fire behaviour. This study proposed, two novel Artificial Intelligence developed models, Multivariate Adaptive Regression Splines (MARS) and Random Forest (RF) methods, to model and predict peak heat release rate (pHRR) of Polymethyl methacrylate (PMMA) from Microscale Combustion Calorimetry (MCC) experiment. From the statistical analysis, MARS presented the highest coefficient of determination (R²) values of (0.9998) and (0.9996) for training and testing respectively, with low MAD, MAPE and RMSE values. Comparatively, MARS outperformed RF in the predictions of pHRR, through its model algorithms that generated optimized equations for pHRR predictions, covering all non-linearity points of the experimental data. Amongst the input variables (sample mass, THR, HRC, pTemp and pTime), heating rate (β), highly influenced pHRR outcome predictions from MARS and RF models. However, to validate the performance and applicability of the proposed models. Results of MARS and RF were benchmarked with that from Artificial Neural Network (ANN) methods. The MARS and RF models observed the least error deviation when compared with pHRR results for PMMA from the ANN models. This study therefore, recommends the adoption of MARS and RF in the predictions of flammability characteristics of polymeric materials.

Data sets in large applications are often too gigantic to fit completely inside the computer’s internal memory. The resulting input/output communication (or I/O) between fast internal memory and slower external memory (such as disks) can be a major performance bottle−neck. While applying sorting on this huge data set, it is essential to do external sorting. This paper is concerned with a new in−place external sorting algorithm. Our proposed algorithm uses the concept of Quick−Sort and Divide−and−Conquer approaches resulting in a faster sorting algorithm avoiding any additional disk space. In addition, we showed that the average time complexity can be reduced compared to the existing external sorting approaches.

The possibility of errors being propagated during the encoding process of cryptographic and steganographic schemes is real due to the introduction of noise by ciphering the data from stage to stage. This real possibility therefore requires that an efficient scheme is proposed such that if after the decoding process the accurate information is not discovered, then it can be employed to detect and correct any errors in the system. The Residue Number System (RNS) by its nature is fault tolerant since an error in one digit position does not affect other digit positions; but the Redundant Residue Number System (RRNS) had been used over the years to effectively detect and correct errors. In this paper, we propose an efficient scheme that can detect and correct both single and multiple errors after and/or during computation and/or transmission provided the redundant moduli are sufficient enough. A theoretical analysis of the performance of the proposed scheme show it will be a better choice for detecting and correcting computational and transmission errors to existing similar state-of-the-art schemes.