Arghasree Banerjee

Bio: Arghasree Banerjee is an academic researcher from University of Engineering & Management. The author has contributed to research in topics: Image segmentation & Encryption. The author has an hindex of 3, co-authored 13 publications receiving 45 citations.

Imbalanced Twitter Sentiment Analysis using Minority Oversampling

Abstract: Micro-Blogging platforms have become one of the popular medium which reflects opinion/sentiment of social events and entities. Machine learning based sentiment analyses have been proven to be successful in finding people’s opinion using redundantly available data. However, current study has pointed out that the data being used to train such machine learning models could be highly imbalanced. In the current study live tweets from Twitter have been used to systematically study the effect of class imbalance problem in sentiment analysis. Minority oversampling method is employed here to manage the imbalanced class problem. Two well-known classifiers Support Vector Machine and Multinomial Naive Bayes have been used for classifying tweets into positive or negative sentiment classes. Results have revealed that minority oversampling based methods can overcome the imbalanced class problem to a greater extent.

Synthetic minority oversampling in addressing imbalanced sarcasm detection in social media

Abstract: Recent developments in sarcasm detection have been emerged as extremely successful tools in Social media opinion mining. With the advent of machine learning tools, accurate detection has been made possible. However, the social media data used to train the machine learning models is often ill suited due to the presence of highly imbalanced classes. In absence of any thorough study on the effect of imbalanced classes in sarcasm detection for social media opinion mining, the current article proposed synthetic minority oversampling based methods to mitigate the issue of imbalanced classes which can severely effect the classifier performance in social media sarcasm detection. In the current study, five different variants of synthetic minority oversampling technique have been used on two different datasets of varying sizes. The trustworthiness is judged by training and testing of six well known classifiers and measuring their performance in terms of test phase confusion matrix based performance measuring metrics. The experimental results indicated that SMOTE and BorderlineSMOTE – 1 are extremely successful in improving the classifier performance. A thorough analysis has been performed to better understand the effect of imbalanced classes in social media sarcasm detection.

Data Security Techniques Based on DNA Encryption

Abstract: Security of the digital data is one of the major concerns of the today’s world. There are several methods for digital data security that can be found in the literature. Biological sequences have some features that make it worthy for the digital data security processes. In this work, DNA encryption and its different approaches are discussed to give a brief overview on the data security methods based on DNA encryption. This work can be highly beneficial for future research on DNA encryption and can be applied on different domains.

Biomedical Image Security Using Matrix Manipulation and DNA Encryption

Abstract: Biomedical image analysis is an integral part of the modern healthcare industry and has a huge impact on the modern world. Automated computer-aided systems are highly beneficial for fast, accurate and efficient diagnosis of the biomedical images. Remote healthcare systems allow doctors and patients to perform their jobs from separate geographic locations. Moreover, expert opinion about a patient can be obtained from a doctor who is in a different country or in some distant location within stipulated amount of time. Remote healthcare systems require digital biomedical images to be transferred over the network. But several security threats are associated with the transmission of the biomedical images. Privacy of the patients must be preserved by keeping the images safe from any unauthorized access. Moreover, the contents of the biomedical images must be preserved efficiently so that no one can tamper it. Data tampering can produce drastic results in many cases. In this work, a method for biomedical image security has been proposed. DNA encryption method is one of the emerging methods in the field of cryptography. A secure and lossless encryption method is developed in this work. Various numerical parameters are used to evaluate the performance of the proposed method which proves the effectiveness of the algorithm.

A Biomedical Image Segmentation Approach Using Fractional Order Darwinian Particle Swarm Optimization and Thresholding

Abstract: Image segmentation is one of the inevitable parts of the digital image processing and very useful to solve different real life problems. Biomedical image segmentation is a prime domain of application of digital image processing. and automated computer aided diagnostics process has high dependency on it. Automated identification of different regions of an image are often required by the human experts. Moreover, accurate detection and identification of a region of interest is possible using the automated methods. Errors are common for the human experts and can be reduced and faster results can be achieved with the help of automated and intelligent systems. This work proposes a biomedical image segmentation process using Fractional Order Darwinian Particle Swarm Optimization (FODPSO) and thresholding. The efficiency of the proposed method is tested both visually and quantitatively and the results speaks itself about the efficiency of the proposed work.

An Overview of Biomedical Image Analysis From the Deep Learning Perspective

Abstract: Biomedical image analysis methods are gradually shifting towards computer-aided solutions from manual investigations to save time and improve the quality of the diagnosis. Deep learning-assisted biomedical image analysis is one of the major and active research areas. Several researchers are working in this domain because deep learning-assisted computer-aided diagnostic solutions are well known for their efficiency. In this chapter, a comprehensive overview of the deep learning-assisted biomedical image analysis methods is presented. This chapter can be helpful for the researchers to understand the recent developments and drawbacks of the present systems. The discussion is made from the perspective of the computer vision, pattern recognition, and artificial intelligence. This chapter can help to get future research directions to exploit the blessings of deep learning techniques for biomedical image analysis.

Feature Selection based on Genetic Algorithm and Hybrid Model for Opinion Mining

Abstract: Sentiment classification is to find the polarity of product or user reviews. Supervised machine learning algorithms is used for opinion mining such as Naive Bayes, K-nearest neighbor, Decision Trees, Maximum Entropy and Hidden Markov Model and Support Vector Machine. KNN is a simple algorithm, but a less efficient classification algorithm. In this paper, we propose an improved KNN algorithm. An optimized feature selection, genetic algorithm that incorporates the information gain for feature reduction and combined with bagging technique. The new method improves the accuracy of sentiment classification. Specifically, we compared two approaches with PCA feature reduction technique and traditional KNN for Sentiment Classification of movie reviews. The same approach has been applied to other machine learning algorithms such as Support Vector Machine and Naive Bayes. The proposed method is evaluated and experimental results using information gain, genetic algorithm with bagging technique indicate higher performance result with accuracy of 87.50% of the movie reviews and exhibits better performance in terms of Accuracy, Precision and Recall for Movie, Book, DVD, Electronics and Kitchen reviews.

Synthetic minority oversampling in addressing imbalanced sarcasm detection in social media

Abstract: Recent developments in sarcasm detection have been emerged as extremely successful tools in Social media opinion mining. With the advent of machine learning tools, accurate detection has been made possible. However, the social media data used to train the machine learning models is often ill suited due to the presence of highly imbalanced classes. In absence of any thorough study on the effect of imbalanced classes in sarcasm detection for social media opinion mining, the current article proposed synthetic minority oversampling based methods to mitigate the issue of imbalanced classes which can severely effect the classifier performance in social media sarcasm detection. In the current study, five different variants of synthetic minority oversampling technique have been used on two different datasets of varying sizes. The trustworthiness is judged by training and testing of six well known classifiers and measuring their performance in terms of test phase confusion matrix based performance measuring metrics. The experimental results indicated that SMOTE and BorderlineSMOTE – 1 are extremely successful in improving the classifier performance. A thorough analysis has been performed to better understand the effect of imbalanced classes in social media sarcasm detection.

An Unsupervised Fuzzy Clustering Approach for Early Screening of COVID-19 From Radiological Images

Abstract: A global pandemic scenario is witnessed worldwide owing to the menace of the rapid outbreak of the deadly COVID-19 virus. To save mankind from this apocalyptic onslaught, it is essential to curb the fast spreading of this dreadful virus. Moreover, the absence of specialized drugs has made the scenario even more badly and thus an early-stage adoption of necessary precautionary measures would provide requisite supportive treatment for its prevention. The prime objective of this article is to use radiological images as a tool to help in early diagnosis. The interval type 2 fuzzy clustering is blended with the concept of superpixels, and metaheuristics to efficiently segment the radiological images. Despite noise sensitivity of watershed-based approach, it is adopted for superpixel computation owing to its simplicity where the noise problem is handled by the important edge information of the gradient image is preserved with the help of morphological opening and closing based reconstruction operations. The traditional objective function of the fuzzy c-means clustering algorithm is modified to incorporate the spatial information from the neighboring superpixel-based local window. The computational overhead associated with the processing of a huge amount of spatial information is reduced by incorporating the concept of superpixels and the optimal clusters are determined by a modified version of the flower pollination algorithm. Although the proposed approach performs well but should not be considered as an alternative to gold standard detection tests of COVID-19. Experimental results are found to be promising enough to deploy this approach for real-life applications.