Journal of Telematics and Informatics
Not a member yet
147 research outputs found
Sort by
Machine Learning Approaches on External Plagiarism Detection
External plagiarism detection is a technique that refers to the comparison between suspicious document and different sources. External plagiarism models are generally preceded by candidate document retrieval and further analysis and then performed to determine the plagiarism occurring. Currently most of the external plagiarism detection is using similarity measurement approaches that are expressed by a pair of sentences or phrase considered similar. Similarity techniques approach is more easily understood using a formula which compares term or token between the two documents. In contrast to the approach of machine learning techniques which refer to the pattern matching and cannot directly comparing token or term between two documents. This paper proposes some machine learning techniques such as k-nearest neighbors (KNN), support vector machine (SVM) and artificial neural network (ANN) for external plagiarism detection and comparing the result with Cosine similarity measurement approach. This paper presented density based that normalized by frequency as the pattern. The result showed that all machine learning approach used in this experiment has better performance in term of accuracy, precision and recall
Detection and Classification of three phase Power Quality events using Wavelets Transforms and Soft Computing Techniques
Analysis of power quality and its related problems is of very much important for both the utilities and end users. There are a large number of concerned authorities to monitor and mitigate the power quality problems. It requires a larger amount to deliver a poor power. So by considering the global economical losses, it is very much urgent to mitigate the various problems affecting the true power. Classification of problems is equally important to mitigation. The common power quality problems occurring are voltage sag, swell, harmonics, flickers etc. Good power determines the fitness of electric power to consumer devices and appliances. It is very important to maintain the detection accuracy of power quality events throughout the operation span. This paper deals with a study based on signal processing algorithms and soft computing techniques for the detection, classification and estimation of power quality events. The literature review points toward the application of wavelet transforms with different filters for achieving feature extraction. The power quality disturbance model is simulated using Simulink toolbox. It is observed that every power quality wavelet disturbance will show unique characteristics and it is generally used to provide an adoptable classification of power quality events. Because of the non-stationary and transitory behavior of the power quality events, the classification goes on challenging and demanding. Thus the feature extraction along with artificial neural network and fuzzy logic incorporated as a powerful too
One-Minute Derivation of The Conjugate Gradient Algorithm
One of the great triumphs in the history of numerical methods was the discovery of the Conjugate Gradient (CG) algorithm. It could solve a symmetric positive-definite system of linear equations of dimension N in exactly N steps. As many practical problems at that time belonged to this category, CG algorithm became rapidly popular. It remains popular even today due to its immense computational power. But despite its amazing computational ability, mathematics of this algorithm is not easy to learn. Lengthy derivations, redundant notations, and over-emphasis on formal presentation make it much difficult for a beginner to master this algorithm. This paper aims to serve as a starting point for such readers. It provides a curt, easy-to-follow but minimalist derivation of the algorithm by keeping the sufficient steps only, maintaining a uniform notation, and focusing entirely on the ease of reader
Average Hashing for Perceptual Image Similarity in Mobile Phone Application
Common problem occurs in almost all mobile devices was duplicated data or files. Such as duplicated images that often happen by event like capturing perceptually similar photos by the user, or images that shared several times in messaging applications chat groups. This common problem can be solved by manually search and remove the duplicated images one by one by the users, but better solutions is by building automated application that search perceptually similar images then provide the result to the users. We study and implementing Average Hashing and Hamming distance for perceptual image similarity into application under mobile phone platform to realize the solution for the problem. The result was very promising in speed and accuracy for finding perceptually similar images under limited resources device like mobile phone
A Mobility Aware Schema To Lower Packet Losses For Reactive Ad Hoc Routing Protocols
In an increasingly wireless world, Ad hoc networks have progressively been developed to become a suitable replacement or, at least complement to the traditional wireless networks. In contrast to the conventional wireless network, the Ad hoc network does not necessarily need a preset infrastructure to be functional. The network members move freely and arrange themselves to perform the routing operations in a distributed fashion. The kernel technology enabling such flexibility is the routing protocol. Though the absence of the infrastructure lowers the installation, configuration and maintenance cost of the dynamic networks, the mobile nodes must deal with several complex issues. Thus, managing : a) security, b) power consumption and balance, c) Quality of Service, in a distributed manner is very challenging. However, above all else, maintaining a path formed by moving particles is the most problematic task. In this paper we analyse the effect of the nodes mobility on the routing protocol performance. Then, we propose a simple schema to counter this issue
Design of Computer Based Signal Generator
The project includes various types of signals alternative study and representation of a computer in digital format and out through the port as an outlet for the output of digital data to generate signals and electronic control circuit of the printer parallel LPTI prevailing used to generate signals. The aim of the research is to evaluate the usage of the computer including the advantage of its possibilities, capabilities in multiple signals processing, and control signal generation methods in many ways that are used to generate various signals at frequencies different capacities. The necessary software requires generating different types of signals, frequencies, and capacities. It is as well as the different software required for the purpose of controlling the electronic circuit. It has been relying on a programmatic method depends on the language of Visual Basic. In addition to the study of the physical system requirements, the analysis how to connect them with the computer and version control signals E-circle this study using speed of a transferring time for generator is Tc = 100nsec. The Speed of a computer which is used to treat ‘1000 MHz’, is got on the farthest frequency to get out from the circle is ‘2500 Hz’. It is used to different signals, sine, triangular or square in 30 samples. The sample numbers are tested. The result of this study shows that each of the sample number, input frequency, and sample time influence the results clearly. If the sample numbers increase, the accuracy of the waves increases. If the input frequency increases, the width of the wave losses. Keywords: Signal Generator, Computer, Visual Basic, Signals
Provision of Relevant Results on web search Based on Browsing History
Different users submit a query to a web search engine with different needs. The general type of search engines follows the "one size fits all" model which is not flexible to individual users resulting in too many answers for the query. In order to overcome this drawback, in this paper, we propose a framework for personalized web search which considers individual's interest introducing intelligence into the traditional web search and producing only relevant pages of user interest. This proposed method is simple and efficient which ensures quality suggestions as well as promises for effective and relevant information retrieval. The framework for personalized web search engine is based on user past browsing history. This context is then used to make the web search more personalized. The results are encouraging
Retail Web System Upgrading with Strategic Customer Using Threshold Policy
This paper mainly deals with the problems that online retailers are facing by some group of strategic customers by upgrading web system and goods pricing. A monopolist price of a product for which stylized dynamic pricing model is considered at the beginning and the price change is done on the product after considering some set of constraints based on the product availability. Due to some network issues sometimes website may get problems at that time there may be a chance of transaction failure for customer during online purchasing. To predict this probability and doing purchasing opinion depending on the belief of Transaction Success Probability (TSP).Considering this there will be a threshold policy for purchasing by customers here it is going to consider it as first one: Customers buy products only if his valuation for products is above threshold (total reduced profit) otherwise they will not take it into consideration. The threshold rises as TSP degrades, or customers turn out to be risk averse. Here we obtain the best cost of each phase and make out the best strategy for online shopping site customizing by threshold and upgrade the system when present TSP is less than threshold. The online trader turns to increase the price if he disregards strategic customer behavior. The value of disregarding strategic customer deeds is significant. The profit loss cost of disregarding strategic customer deeds grows as customer transaction cost rises
Significance of Weighted-Type Fractional Fourier Transform in FIR Filters
The desired frequency response of a filter is periodic in frequency and can be expanded in Fourier series. One possible way of obtaining FIR filter is to truncate the infinite Fourier series. But abrupt truncation of the Fourier series results in oscillation in the pass band and stop band. These oscillations are due to slow convergence of the Fourier series by the Gibb’s phenomenon. To reduce these oscillations the Fourier coefficients of the filter are modified by multiplying the infinite impulse response with a finite weighing sequence called a window. The Fourier transform (FT) of a window consists of a central lobe and side lobes. The central lobe contains most of the energy of the window. To get an FIR filter, the desired impulse response and window function are multiplied, which results to give finite length non-causal sequence. Since Fractional Fourier Transform (FrFT) is generalization of FT. Here an attempt is to implement filters using window by using Weighted Type Fractional Fourier Transform (WFrFt), differentiator and integrator using weighted FrFt is also present
Automated Home Appliances Control using Embedded Web Server
In modern construction in industrialized nations, most homes have been wired for TV, lights and fans etc... Many household tasks were automated by the development of specialized automated appliances. In this project we are going to create the embedded web server so that we can control the electrical devices through online by using either computer or by using smart phone. The purpose of this project is to design a control system that able to control a system device remotely from distance and monitoring condition of the system in real-time as well. This project shows how to set up an Arduino with Ethernet shield as a web server. The web servers in this Project are used to serve up web pages that can be accessed from a web browser and allows our hardware to be controlled from web browser. Starting from basic, our example for this project is simple. We want to access our device by switching ON-OFF LED and monitor analog input data acquisition from web server, which are the computer and Arduino connected to the same network