T. Vetriselvi

Work place: Department of CSE, K. Ramakrishnan College of Technology, Tiruchirappalli-621112, India

E-mail: vetriselvi09@gmail.com


Research Interests: Data Mining, Programming Language Theory


T. Vetriselvi: Part-Time Research Scholar at Department of Computer Applications, National Institute of Technology Tiruchirappalli. She is currently an assistant professor at the K. Ramakrishnan College of Technology, Tiruchirappalli. She has 11 years of teaching experience. Her areas of interest include Data Mining and Programming Languages.

Author Articles
Key Term Extraction using a Sentence based Weighted TF-IDF Algorithm

By T. Vetriselvi N.P.Gopalan G. Kumaresan

DOI: https://doi.org/10.5815/ijeme.2019.04.02, Pub. Date: 8 Jul. 2019

Keyword ranking with similarity identification is an approach to find the significant Keywords in a corpus using a Variant Term Frequency Inverse Document Frequency (VTF-IDF) algorithm. Some of these may have same similarity and they get reduced to a single term when WordNet is used. The proposed approach that does not require  any test or training set, assigns sentence  based Weightage to the keywords(terms) and it  is found to be  effective. Its suitability is analyzed with several data sets using precision and recall as metrics.

[...] Read more.
An Efficient Image Block Encryption for Key Generation using Non-Uniform Cellular Automata

By G. Kumaresan N.P.Gopalan T. Vetriselvi

DOI: https://doi.org/10.5815/ijcnis.2019.02.04, Pub. Date: 8 Feb. 2019

Cryptographic image block encryption schemes play a significant role in information enabled services. This paper proposes an image block encryption scheme based on a novel three stage selection (TSS) method in a public cloud with reversible cellular automata. Due to the openness of public cloud, different attacks are possible over user sensitive information. The TSS method has three stages and they generate a robust master key with user plaintext as input and produces an encrypted block as key to be sent to authenticated users. An analysis of experimental results shows that this new method has a large key space and immune to brute force attacks, statistical cryptanalysis attacks and chosen plaintext attacks. Also, the encrypted image entropy value could be increased to 7.9988 making it ideal for a best image block encryption for key generation.

[...] Read more.
Other Articles