Manuscripts

Recent Papers

Research Paper

An implementation framework for real-time spam detection in Twitter

With the expanded prominence of online informal community, spammers discover these stages effectively available to trap clients in noxious exercises by posting spam messages. In this work, we have taken the Twitter stage and performed spam tweets identification. To stop spammers, Google Safe Perusing and Twitter's BotMaker instruments identify and square spam tweets. These instruments can square noxious connections, anyway, they can't ensure the client continuously as ahead of schedule as could be expected under the circumstances. Along these lines, businesses and specialists have connected diverse ways to deal with make spam free informal community stage. Some of them are just founded on client-based highlights while others depend on tweet based highlights as it were. Nonetheless, there is no extensive arrangement that can solidify tweet's content data alongside the client based highlights. To illuminate this issue, we proposed a system which takes the client and tweet based highlights alongside the tweet content component to order the tweets. The advantage of utilizing tweet content element is that we can recognize the spam tweets regardless of whether the spammer makes another record which was unrealistic just with client and tweet based highlights. We have evaluated our solution with two different machine learning algorithms namely – Support Vector Machine and Random Forest. We are able to achieve an accuracy of 86.75% and surpassed the existing solution by approximately 17%.

Published by: Poornima N. C.

Author: Poornima N. C.

Paper ID: V5I2-2050

Paper Status: published

Published: April 23, 2019

Full Details
Research Paper

A secure client side deduplication scheme in cloud storage environments

Freshly identified and endorsed the trend of holding cloud-based services for large scale content storage, processing, and distribution. Security is the top concerns for open cloud environments. Towards these security challenges, we propose and implement, on Open Stack Swift, a new client-side deduplication scheme for securely storing and sharing outsourced data via the public cloud. The originality of our proposal is twofold. First, it ensures better confidentiality towards unauthorized users. That is, every client computes a per data key to encrypt the data that he intends to store in the cloud. hence, data access is with the data owner. Second, by accessing rights in the metadata file, an authorized user can decipher an encrypted file only with his private key. It elaborates a computationally low-cost method for making all log entries generated prior to the logging machine's compromise impossible for the attacker to make an attempt and also impossible to undetectably modify or destroy.

Published by: Prateek Naharia, Eshan Pacheriwala, Soumyajyoti Dutta, M. Vinodhini

Author: Prateek Naharia

Paper ID: V5I2-2049

Paper Status: published

Published: April 23, 2019

Full Details
Review Paper

Noise reduction in electrocardiogram using LMS algorithm based on adaptive filtering

This project tells about the algorithms of adaptive filters which is for reduction of noise in these heat waves. There are different types of algorithms like NLMS, LMS, and RLS. In those types there are only LMS is used mostly for noise reducing because it produces low MMSE than all of those algorithms so for the best accuracy the LMS algorithm is used. By using mat lab we calculate the performance. The measures we calculated in mat lab is in range of convergence value and MMSE value. When we decrease the step size of noise will also decrease. If it is increased convergence value to increases, then the rate of noise also will increase.

Published by: Jayaprakash A., Manikanta Sai G., Satya Praveen P., Vishnu Teja I., Mohammed Arif A.

Author: Jayaprakash A.

Paper ID: V5I2-2012

Paper Status: published

Published: April 23, 2019

Full Details
Research Paper

Extensions of Kannan Fixed-Point theorem on complete metric spaces and on generalized metric spaces

In this paper we get sufficient conditions for the existence of unique fixed point of Kannan type mappings on complete metric spaces and on generalized complete metric spaces depended on another function.

Published by: Dharmendra Singh Rajput, Anil Rajput, Rucha Athaley

Author: Dharmendra Singh Rajput

Paper ID: V5I2-2007

Paper Status: published

Published: April 23, 2019

Full Details
Research Paper

Identification of currency denomination using image processing

This paper aims to solve the problem of detecting a paper currency and output the result as speech. The developed model accepts an image from a webcam (RasPi Camera). The Raspberry Pi’s image processing algorithm then extracts certain features from this image and matches them against a set of training data. When an appropriate match of the acceptable threshold is found, the device outputs the denomination value through speech synthesis. Moreover, this paper also compares two methods of image processing namely, Brute Force Matcher Algorithm and Convolution Neural Network Algorithm. Finally, the pros and cons of both the methods are evaluated and the best algorithm is ascertained.

Published by: Gokul Ramasamy, Sakthi Subramanian

Author: Gokul Ramasamy

Paper ID: V5I2-2002

Paper Status: published

Published: April 23, 2019

Full Details
Research Paper

Correlation between the water level and the growth of weeds in the paddy field in Kannur District, Kerala

The presence of water in the field appears to inhibit the growth of weeds, which are largely terrestrial or semi-aquatic species. The lack of retention of water in the Midland Laterite Hillock Paddy Field (MLHPF) favored the excess growth of weeds. But in the other two regions (ie FWPF and CPF), the weed growth was less because these regions were filled with water up to a certain extent during the southwest monsoon.

Published by: Prabhakaran P. V., Dr. Aravindan Tharemmal

Author: Prabhakaran P. V.

Paper ID: V5I2-1998

Paper Status: published

Published: April 23, 2019

Full Details