Shanmuga Priya S currently serves as Assistant Professor at the Department of Computer Science and Engineering, School of Engineering, Amrita Vishwa Vidyapeetham, Coimbatore. She received her M. E. degree in Computer Science and Engineering from Anna University. Her areas of interest include Programming, Data Structures and Algorithms, Computer Networks. Her areas of research include Pattern Recognition, Time Series Analytics.


Publication Type: Conference Paper

Year of Publication Title


S. Prabhu Thandapani, Senthilkumar, S., and Shanmuga Priya S., “Decision Support System for Plant Disease Identification”, in International Conference on Advanced Informatics for Computing Research. Springer, Singapore, 2019.[Abstract]

The significance of agriculture in India and the amount of damage to the sector due to plant diseases, calls for a system which can identify plant diseases accurately. Improper identification of diseases and taking wrong measures to prevent the disease will be cost inefficient and time consuming. There are highly accurate existing techniques to identify plant diseases but they are specific to a particular crop. In this project, a generic system is developed to identify plant diseases accurately based on textual description of plant diseases. Dataset containing description of diseases is created using the concept of web scraping. The dataset is preprocessed where, keywords are extracted, categorized and grouped to obtain the list of symptoms from the disease description. Based on the symptoms provided by the user to the system, plant disease is identified and the output is the identified disease along with preventive measures.

More »»

Publication Type: Journal Article

Year of Publication Title


Shanmuga Priya S. and Abinaya, M., “Feature selection using random forest technique for the prediction of pest attack in cotton crops.”, International Journal of Pure and Applied Mathematics, vol. 118, pp. 2899-2902, 2018.[Abstract]

Incorporating the technology into the Agriculture field for triggering the growth and identifying the pest/disease in various crops is the state of art. Extreme Changes in the weather parameters during the growth phase of crops causes serious threats which tends to go in search for new adaptive measures. The main aim of this work is to predict the occurrence of the pest in cotton crop based on such weather factors using clustering techniques. A change in any of these meteorological factors fluctuates the infestation of the pest between higher and lower. The values of the weather parameters computed which cause the occurrence of these pests can be used for prediction and precautionary measures can be taken ahead. © 2017 Academic Press. All Rights Reserved.

More »»


Shanmuga Priya S. and Dr. Padmavathi S., “Comparative analysis of classification algorithms for predicting the advertisements on webpages”, Asian Journal of Information Technology, vol. 15, no. 4, pp. 738-742, 2016.[Abstract]

The fast growth of the Internet has completely changed the way people using computers. In the current scenario, people are more exposed to media and internet has led to the creation of advertisement which can reach users and it has become the ultimate for most business to enhance their profit. More and more ads are being sold on a single-impression basis as opposed to bulk purchases. Identifying whether an image belongs to advertisement or not is of interest to many internet users. This study analyses the performance of probabilistic, tree based and rule based classifier for this classification. Their performances under various conditions are summarized. © Medwell Journals, 2016.

More »»


B. .A.Sabarish and Shanmuga Priya S., “Improved Data Discrimination in Wireless Sensor Networks”, Scientific research, vol. 4, pp. 117-119, 2012.[Abstract]

In Wireless Sensors Networks, the computational power and storage capacity is limited. Wireless Sensor Networks are operated in low power batteries, mostly not rechargeable. The amount of data processed is incremental in nature, due to deployment of various applications in Wireless Sensor Networks, thereby leading to high power consumption in the network. For effectively processing the data and reducing the power consumption the discrimination of noisy, redundant and outlier data has to be performed. In this paper we focus on data discrimination done at node and cluster level employing Data Mining Techniques. We propose an algorithm to collect data values both at node and cluster level and finding the principal component using PCA techniques and removing outliers resulting in error free data. Finally a comparison is made with the Statistical and Bucket-width outlier detection algorithm where the efficiency is improved to an extent.

More »»