Skip to main content

2023 | Book

Smart Trends in Computing and Communications

Proceedings of SmartCom 2022

Editors: Yu-Dong Zhang, Tomonobu Senjyu, Chakchai So-In, Amit Joshi

Publisher: Springer Nature Singapore

Book Series: Lecture Notes in Networks and Systems


About this book

This book gathers high-quality papers presented at the Sixth International Conference on Smart Trends in Computing and Communications (SmartCom 2022), organized by Global Knowledge Research Foundation (GR Foundation) in partnership with IFIP InterYIT during January 11–12, 2022. It covers the state of the art and emerging topics in information, computer communications, and effective strategies for their use in engineering and managerial applications. It also explores and discusses the latest technological advances in, and future directions for, information and knowledge computing and its applications.

Table of Contents

Implementation of Tweet Stream Summarization for Malicious Tweet Detection

In contrast to traditional media, social media is populated by anonymous individuals who have the freedom to broadcast whatever they choose. This online social media culture is dynamic in nature, and the move from traditional media to digital media is growing increasingly popular among people. While conventional media will continue to be used less frequently in the future, the increasing use of online social networks (OSNs) will obfuscate the real information provided by traditional media. Genuine users provide material that is beneficial to the broader public; on the other hand, spammers transmit irrelevant or misleading content that turns social media into a front for spreading false information. Consequently, undesirable text or susceptible links might be delivered to targeted users. These fabricated texts are anonymous, and they are occasionally linked to possible URLs. A precise statistical classification for a piece of news is not possible with the current systems because of data limitations and communication types. We will look at a variety of research publications that employ a variety of strategies for master training in the prediction and detection of harmful material on social media websites and networks. In this study, we attempted to identify spam tweets from a large collection of tweets by utilizing TVC algorithm to identify it. When a content paper is incorporated in this manner, a brief summary produced by utilizing the most important keywords from the first document is known as summarization. It is necessary to have a dynamic approach of dealing with the condensed information that is supplied through Twitter feeds. This research presents a novel way for producing a major substance-based summary in a shorter period of time than previously available. We also propose to detect harmful tweets both offline and online and to do so in real time. Most notably, when compared to the other current frameworks, the proposed framework performs multi-subject summarization on an online dataset, which results in a reduction in the amount of time required.

Pranoti Mane, Shrikant Sonekar, Sayema Kausar
Comparative Analysis of a Secure Authentication Protocol for 5G-Enabled IoT Network Using Public and Shared Secret Key

Due to advancements in technologies and to achieve high efficiency, the world is shifting from 4G to a 5G network. The advantages of 5G network are enormous in every field and thus are catching speed to be a better version of what exists in today's world. The features like increased bandwidth availability, massive network capacity, and ultra-low latency are keeping 5G out of the box and giving raw network engineers to explore and invent high standards. As the world goes with pros and cons hand in hand, so is the advent of 5G technology. Finding its applications in communication, massive IoT, and mobile broadband, it also faces some challenges which affects the day-to-day life of humans. 5G applications face security risks due to the new technology used and the performance requirements of the specific application scenario. To avail the 5G services, the IoT devices need to communicate with an intermediate network known as an access network, which is usually publicly accessible and thus prone to attacks. In this paper, implementation and results of a 5G-enabled secure authentication protocol using symmetric and asymmetric key are compared to identify the best technique for privacy preservation in 5G-enabled IoT network.

Shamshuddin K, Vishal B. Pattanashetty
DDoS Botnet Attack Detection in IoT Devices

Distributed denial of service (DDoS) attacks are most harmful threats on present Internet. The Internet of things (IoT) weaknesses give an optimal objective to botnets, making them a significant supporter of the expanded number of DDoS attacks. The expansion in DDoS attacks made it essential to address the results as it suggests toward the IoT business as one of the significant causes. The main motive of this paper is to provide an analysis of the attempts to prevent DDoS attacks, which mainly at the network level. The reasonableness of these arrangements is removed from their effect in settling IoT vulnerabilities. It is obvious from this survey that there is no ideal arrangement yet for IoT security, but this field has large chances for innovative work.

Bandari Pranay Kumar, Gautham Rampalli, Pille Kamakshi, T. Senthil Murugan
Text/Sign Board Reading Aid for Visually Challenged People

The technological advancements in today’s world are focussing on designing smart devices that improve lifestyle of human beings by making things easier and simple. Based on this, a productive approach is proposed for detecting the text from the natural scenes and converting it into voice output. The aim of this project is to overcome the reading problems of visually impaired people and their complete dependence on Braille code. The project is developed using MATLAB software in which the algorithms used are maximally stable extremely regions (MSER), stroke width transform, optical character recognition (OCR), and speech synthesizer of MATLAB. By utilizing these techniques, the text from the images can be identified effectively, and it can be converted into speech output.

Srikhakollu Pandu Ranga Avinash, V. Krishna Sree, M. Saahith Reddy, K. Sravani Kumari, A. Gaurav
Automatic Segmentation of Red Blood Cells from Microscopic Blood Smear Images Using Image Processing Techniques

Navya, K. T. Das, Subhraneil Prasad, KeerthanaHuman blood is a very effective parameter to detect, diagnose and rectify ailments of the human body. Complete blood count (CBC) is a method to clinically obtain a statistical measure of blood and its related parameters, i.e., red blood cells (RBCs), white blood cells (WBCs), platelets, hemoglobin concentration to name a few. This helps to determine the physical state of the subject. For further diagnosis, peripheral blood smear, a thin layer of blood smeared on a microscope slide and stained using various staining methods is examined for the morphology of the cells by the pathologists. However, manual inspection of smear images is tedious, time-consuming, and laboratorian-dependent. Although there are certain software-based approaches to tackle the problem, most of them are not robust for all staining methods. Thus, the need is to create an automated algorithm that will work for different staining types, thereby alleviating both the aforementioned drawbacks. This work aims to create an automatic method of segmenting and counting RBCs from blood smear images using image processing techniques to help diagnose RBC-related disorders. In the proposed method, the images are first preprocessed, i.e., standardized to a uniform color and illumination profile using contrast enhancement, adaptive histogram equalization followed by Reinhard stain normalization algorithms. WBCs and platelets are extracted in HSI color space and subtracted from the original image to retain only RBCs. Thereafter using morphological operations and active contour segmentation algorithms, a count of total RBCs were obtained even for overlapped cells in the microscopic blood smear image. The proposed method achieved counting accuracy of 89.6% for 150 images.

K. T. Navya, Subhraneil Das, Keerthana Prasad
Design and Simulation of Microstrip Bandpass Filter Techniques at X-Band for RADAR Applications

A new approach to compact bandpass filter techniques such as end-coupled, parallel-coupled, hairpin, interdigital, and combline filters with their enhancement in size and frequency performance is introduced in this paper. Designed fifth-order filters that are simulated in the compact narrowband frequency is targeted for high selectivity, better return loss, and insertion loss with a passband ripple using Chebyshev response. To minimize the dimension of filters, we implemented an optimized design by determining its parameters, coupling resonators, and connecting elements. The results are simulated using Genesys Keysight software and iterated for the desired specimen. The prototype filter consists of Rogers RT/duroid 5880 substrate with a thickness of 0.508 mm. Experimental result of hairpin filter structure shows a reduced design (20.2 mm × 10 mm) and improvises the stopband characteristics with better harmonic suppression and narrower frequency bandwidth than the other conventional filter for future development in RADAR applications.

Shraddha S. Shinde, D. P. Rathod, ArunKumar Heddallikar
A Novel Approach to Detect Plant Disease Using DenseNet-121 Neural Network

The disease of crops is a major risk to food security and can incur a makeable loss to the people. But, the latest development in deep learning for solving this problem surpasses all the traditional methods in terms of efficiency, time period for detection and accuracy. In this paper, we came up with a rapid identification of leaf image and classify the image to correct class by using classical deep neural network architecture, DenseNet-121. This deep learning model has the ability to recognize 15 types of different plant disease, three of which are healthy ones, for better accurate results. The algorithm is highly optimized to produce results in less than 5 s after being fed into the system. The model’s total testing accuracy for plant disease detection is 99%.

Nilesh Dubey, Esha Bhagat, Swapnil Rana, Kirtan Pathak
Offline Handwritten Signature Forgery Verification Using Deep Learning Methods

Offline signature verification is one of the most challenging tasks in biometric authentication. Despite recent advances in this field using image recognition and deep learning, there are many remaining things to be explored. The most recent technique, which is Siamese convolutional neural network, has been used a lot in this field and has achieved great results. This paper presents an architecture that combines the power of Siamese Triplet CNN and a fully-connected neural network for binary classification to automatically verify genuine and forgery signatures even if the forged signature is highly skilled. On the challenging public dataset for signature verification BHSig260, the proposed model can achieve a low False Acceptance Rate = 13.66, which is slightly better than the reference model. Based on this approach, the one-shot learning should make it possible to determine if the input image is genuine or fraudulent just from one base image. Therefore, our model is expected to be extremely suitable for practical problems, such as banking systems or mobile authentication applications, in which the amount of data for each identity is limited in quantity and variety.

Phan Duy Hung, Pham Son Bach, Bui Trong Vinh, Nguyen Huu Tien, Vu Thu Diep
Potential Applications of Advanced Control System Strategies in a Process industry—A Review

Present paper attempts to review the applications of advanced control strategies based on artificial intelligence techniques and its hybrid counterparts applicable in process industry. This chemical process industry may be textile, paper, water purification plant, sugar mill, leather, steel, or any sub-process which may be common in all these industries. It covers an exhaustive literature review.

Pradeep Kumar Juneja, Sandeep Kumar Sunori, Kavita Ajay Joshi, Shweta Arora, Somesh Sharma, Prakash Garia, Sudhanshu Maurya, Amit Mittal
Analysis of Algorithms for Effective Cryptography for Enhancement of IoT Security

The Internet of Things has emerged as one of the most prevalent technologies of the current times, finding its place in a myriad of applications and is widely used for digitization and automation applications such as smart city development, automated monitoring systems, healthcare, energy management and much more. With more devices being connected to the Internet, one of the biggest challenges faced by the Internet of Things surfaces—privacy and security risks. The vulnerability of IoT devices and networks have been brought to light, presenting a threat to the integrity of data. Cryptography has proved itself as a method to secure communication channels and data, as a way to ensure IoT security. In this paper, we aim to compare cryptographic algorithms, namely—AES, DES, RSA and lightweight cryptographic algorithm Fernet, to determine which cryptographic algorithm is the most efficient and secure, and can thereby minimize the risk to data integrity and security in IoT applications.

Valerie David, Harini Ragu, Vemu Nikhil, P. Sasikumar
Relevance of Artificial Intelligence in the Hospitality and Tourism Industry

The authors provide a crisp yet in-depth summary of the relevance of artificial intelligence in the hospitality and tourism industry. Focusing on artificial intelligence, the chapter draws the attention of the reader toward its usage and role in the lives of the customers and service providers in the hospitality sector. It also focuses on the costs involved in the incorporation of artificial intelligence in the hospitality industry. Some of the salient features of artificial intelligence in the hospitality sector have been discussed in the chapter by elucidating the benefits of artificial intelligence like financial benefits, technological benefits, and resource management benefits. This chapter sought to project the various factors affecting the role of artificial intelligence in the industry and how the hospitality and tourism industry adopts the technology with time, demand, and expectations of the customer. Both perspectives have been highlighted, i.e., from the customer’s point of view as well as the managerial insights. The overall intention and thrust were to provide the reader with perspectives of the importance of advancing stages of artificial intelligence in the chosen industry along with a few contrasting thoughts as artificial intelligence cannot replace the essence of human touch completely as it is one of a kind of a servicescape involving humans as the most crucial and critical link.

Smrutirekha, Priti Ranjan Sahoo, Ravi Shankar Jha
Deep Learning-Based Smart Surveillance System

The role of CCTV cameras has been overgrown in this generation. CCTV cameras are installed all over the places for surveillance and security. Many surveillance systems still require human supervision. Recent advances in computer vision are, thus, seen as an important trend in video surveillance that could lead to dramatic efficiency gains. Various public places like shopping malls, supermarkets, ATMs, banks, and other places, where CCTV cameras are available, are the places we should concentrate on. Security can be characterized in various terms in various settings like robbery distinguishing proof, brutality recognition, odds of a blast, and so on. In jam-packed public places, the term security covers practically a wide range of strange occasions. So, it is important and challenging to build a model which detects these abnormal activities and generates some kind of alert. We used a combination of convolutional neural networks (CNNs) and long short-term memory (LSTM) which involve the concept of deep neural networks. It extracts the spatial–temporal features of the images and calculates the Euclidean distance between the original and reconstructed batch of images. We converted the training videos into images to train the model and calculated the loss between the images to identify the abnormality. To validate the proposed algorithm, 4 datasets as HOLLYWOOD, UCF101, HMDB51, and WEIZMANN are used for action recognition. The proposed technique performs better than the existing one. We made use of Jupiter notebook and Python frameworks.

G. Sreenivasulu, N. Thulasi Chitra, S. Viswanadha Raju, Venu Madhav Kuthadi
Pronunciation-Based Language Processing Tool for Tai Ahom

Words with similar pronunciation may not be of same in its written form. Such similarities can be measured by means of computer algorithms. For English language, an algorithm is available, which may be utilized as a database function “Soundex”. This algorithm returns codes on the base of pronunciation of words. Same code signifies similar pronunciation, and dissimilar code indicates different pronunciation. This paper focuses on development of Soundex for Tai Ahom language. In some cases, a single word carries several different meanings. For such matters, exact meaning may be guessed on the basis of other words in the sentence. For this purpose, we have focused on developing another algorithm backed by sentiment analysis.

Dhrubajyoti Baruah, Joydip Sarmah, Kamal Gayan, Satya Ranjan Phukon
Improving Ecommerce Performance by Dynamically Predicting the Purchased Items Using FUP Incremental Algorithm

Data mining or knowledge discovery is the way toward examining data according to substitute perspectives and summarizing it into accommodating information. This information can be then used to fabricate a pay, decreases costs, or both. Programming made with web mining as its key subject ought to permit clients to isolate information from a wide extent of assessments or centers, demand it, and sum up the affiliations perceived. Taking everything into account, information mining is the way toward discovering affiliations or models among many fields in colossal social instructive assortments. This paper effectively tracks down the rehashed bought things by clients. This proposed algorithm is having a higher running time than the existing FUP incremental algorithm. This algorithm efficiently finds the frequent items, and dynamically the items can be added. The entire history of the frequent item database was added and put into separate clusters. At last, we compare and choose the best-purchased items of the customer and also predict the past purchased items in the history. Based on the output, we can easily find the current status of the customer purchase.

K. Kalaiselvi, K. Deepa Thilak, S. Saranya, T. Rajeshkumar, M. Malathi, M. Vijay Anand, K. Kumaresan
A Novel Approach to Privacy Preservation on E-Healthcare Data in a Cloud Environment

Cloud environment enables healthcare professionals to work together. Despite its many benefits, it faces several difficulties: technical, legal, and managerial. Cloud computing can enhance healthcare facilities by utilizing its skills to assist information transfer within the health monitoring system, and without any geographical limitations, users such as patients, physicians, pharmacists, and health insurance agents can obtain health-associated information at any moment. With the widespread use of healthcare information and communication technology (ICT), creating a stable and sustainable data sharing scenario has attracted increasing interest in both academic research and the healthcare sector. This paper evaluates and compare the present situation with the security criteria for cloud-based medical e-health record. In addition, this paper outlines proposed cloud model and comparison of our proposed cloud model with different frameworks listed in different research papers are discussed through which how we can safeguard patient e-healthcare record over cloud environment.

Kirtirajsinh Zala, Madhu Shukla
Temporal Reasoning of English Text Documents

Temporal expression recognition and normalization (TERN) is promising research area in the field of natural language processing. Automatic temporal information extraction system is in great demand due to rapid growth of digitized information on the Internet. Lot of research has been done in temporal information extraction. But due to variety of temporal expression present in the textual data, interpretation needs more efforts. It is important to understand nature of expression, and then, its interpretation ways can be defined. For example, “on the same day,” it requires information about previous sentences in the text to define reference time rather than considering document creation time as reference time statically. In this paper, a normalization system is proposed that dynamically chose the reference time from text rather than relying on traditional approach of selecting a document creation date as reference time. The system is evaluated on gold standard dataset like WikiWar, AQUAINT, and TempEval-2 datasets.

Parul Patel
Design, Development, and Integration of DMA Controller for Open-Power-Based Processor SOC

While the processor is momentarily disabled or busy performing other orders in parallel, the Direct Memory Access (DMA) technology allows direct access to peripherals and memory. DMA gets control of the buses to transfer the data directly to the I/O devices. DMA completes the data transfer to all peripherals without the interference of the processor. The DMA controller supports eight channels with 32-bit data transfer, and it has an interface toward user logic for data read and write. The channel assignment is done based on priority. This project proposes to design, develop, and integrate DMA controller for open-power processor A2O core-based fabless SoC through AXI4 interface. The methodology used for designing is as follows: design state machines, develop Verilog HDL code, simulate using ModelSim Questa®, and synthesis using Vivado Design Suite-Xilinx®.

Gannera Mamatha, M. N. Giriprasad, Sharan Kumar
A Comparative Study Using Numerical Simulations for Designing an Effective Cooling System for Electric Vehicle Batteries

As technology is advancing rapidly in electric vehicle (EV), numerous automobile industries are focusing on shifting a part of their catalogue or have already started production of EVs. One of the most significant parts of an electric vehicle is battery which powers the vehicle, and it usually contains thousands of cells, which can be arranged in various combinations of series and parallel arrangements to form a module, many such modules together make a battery. Lithium-ion batteries are one of the preferred types for use in electric vehicles. The efficiency and life cycle of a lithium-ion battery depend on several factors. Battery temperature is one such critical factor and needs to be monitored to avoid early failure of the battery. Therefore, advanced cooling technology is to be incorporated into a battery to lower the temperature. This paper presents a comparative study of direct and tube cooling methods in a battery for effective cooling of an EV using ANSYS Fluent. These cooling systems are designed keeping an easier process of manufacturing in mind.

Akshat Maheshwari, Vineet Singhal, Ajay Kumar Sood, Meenakshi Agarwal
Lexical Resource Creation and Evaluation: Sentiment Analysis in Marathi

In India, the raise of regional language contents over social media, websites, blogs and news article are exponentially increasing because of ease of use of technology, and people are expressing their thoughts, opinion more conveniently and powerfully over Internet. In this paper, we evaluate the challenges of sentiment analysis in Marathi by setting up a baseline, where we produced an annotated dataset, however, initially, we created an annotated dataset consisting of Marathi news scraped from various newspaper/channel websites. Furthermore, domain experts annotated Marathi news with positive, negative and neutral polarity. And we used machine learning models such as logistic regression, Stochastic Gradient Decent (SGD), support vector machine (SVM), nearest neighbour, neural network, decision tree (DT), Naïve Bayes (NB) and proposed ensemble-based model for sentiment analysis to demonstrate effectiveness. In experimentation, the proposed ensemble classifier outperforms other classifiers with an accuracy of 94.16% and an F-score of 97.02% for fivefold validation. Also, for tenfold validation, the accuracy is 95.07%, and the F-score is 96.93%.

Mahesh B. Shelke, Saleh Nagi Alsubari, D. S. Panchal, Sachin N. Deshmukh
Artificial Neural Network Method for Appraising the Nephrotic Disease

In the shape of an affected person’s proof, the medical report is an ever-developing supply of record for a medical institution. One of the complex issues that get up in the transplanted kidneys is glomerulonephritis. In AI, there are two methodologies: managed and solo mastering. Characterization is a method that falls underneath controlled learning. Out of numerous arrangement models, the maximum prevalently applied is the artificial neural community. While neural networks turn out tremendous in characterization and preparing a device, the precision of the outcome may also anyways be beneath inquiry. The enhancement of the artificial neural networks is completed by using the exactness and space of the result. For this, ANN may be hybridized with a metaheuristic algorithm referred to as the cat swarm optimization (CSO) set of rules. The benefits of optimization artificial neural community are normally the development in the precision of the order, translation of the statistics, and reduction in fee and time utilization for buying real outcomes and so forth within the prevailing study, a correlation between the aftereffects of an ANN decrease again propagation version and the proposed ANN-CSO version is carried out for medical assessment.

K. Padmavathi, A. V. Senthilkumar, Ismail Bin Musirin, Bınod Kumar
Artificial Neural Network and Math Behind It

Article is structured in such a way that the reader could easily understand from the roots of an artificial neuron to its applications in between gaining knowledge on how an artificial neuron exists. It starts with the history of neuron, and the explanation goes on by describing the architecture of artificial neuron and on to the functioning of neuron. It also explains the basic math involved in constructing an artificial neuron by describing the activation functions used in the network, and the mathematical models helps in building the neural network. It also talks about the structure of feed forward network, loss function imposing an error on the output, also the gradient descent, and the back propagation method involving the computation of cost function with respect to the network parameters. As a conclusion, paper describes the importance of artificial neural networks in the upcoming artificial intelligence and many more applications behind it.

Harshini Pothina, K. V. Nagaraja
Machine Learning-Based Multi-temporal Image Classification Using Object-Based Image Analysis and Supervised Classification

During last decade, there has been tremendous research related to the image-based technique in remote sensing; object-based classification is one of the popular techniques due to its capacity of promising results. This paper presents a novel approach where a hybrid method of object-based image analysis and supervised classification is used. The data used in this study is high-resolution multispectral 4-band images from 2017 to 2019 provided by the PlanetScope satellite of region Chandigarh, India. First, the data has been pre-processed through passing it in a pipeline of steps followed by a multi-resolution segmentation algorithm and classifying the image into seven classes based on the spectral signature using algorithms like maximum likelihood (ML), support vector machine (SVM), Mahalanobis distance (MD). Comparing the three algorithms, it was observed that SVM and ML have given the highest overall accuracy of 95.21% and kappa coefficient = 0.9159. Also, the overall accuracy 91.91% and kappa coefficient = 0.8860 were achieved.

Swasti Patel, Priya Swaminarayan, Simranjitsingh Pabla, Mandeepsingh Mandla, Hardik Narendra
A Model to Generate Benchmark Network with Community Structure

Meena, Shyam Sundar Tokekar, VrindaThe studies of social networks focus on the structure and components of networks at different levels. To identify the component in a network, researchers have developed various community detection algorithms. To test the quality of community detection results, networks with well-known community structures are used. But, a very few networks are available for this purpose. Researchers have suggested some models that generate artificial networks with the community. However, most of the proposed models are unable to produce benchmark networks similar to the real-world network. We propose a model that generates benchmark networks for the evaluation of community detection algorithms. The proposed model has been compared with well-known LFR Lancichinetti et al. (Phys Rev 78(4):046110, 2008 [14]) and GLFR Le et al. (2017 26th international conference on computer communication and networks (ICCCN). IEEE, pp 1–9, 2017[15]) models. For performance testing, various structural properties have been analyzed, which are followed by real-world networks. The NMI scores achieved by well-known community detection algorithms were also compared. In experimental analysis, we found that networks generated by our model follow essential properties of real-world networks.

Shyam Sundar Meena, Vrinda Tokekar
Human Activity Recognition Using LSTM with Feature Extraction Through CNN

Human activity recognition is important for detecting anomalies from videos. The analysis of auspicious activities using videos is increasingly important for security, surveillance, and personal archiving. This research paper has given a model which can recognize activities in random videos. The architecture has been designed by using BiLSTM layer which helps to learn a system based on time dependencies. To convert every frame into a featured vector, the pre-trained GoogLeNet network has been used. The evaluation has been done by using a public HMDB51 data set. The accuracy achieved by using the model is 93.04% for ten classes and 63.96% for 51 classes from same data set only. Then, this network is compared with other state-of-the-art method, and it proves to be a better approach for the recognition of activities. Abstract should summarize the contents of the paper in short terms, i.e. 150–250 words.

Rosepreet Kaur Bhogal, V. Devendran
Sentiment Classification of Higher Education Reviews to Analyze Students’ Engagement and Psychology Interventions Using Deep Learning Techniques

Globally, higher education institutions are closed due to the COVID-19 pandemic. The sudden shift to online education excites most teachers and students. The professors are researching online learning platforms. They are only involved in face-to-face teaching in traditional teaching platforms. There are many concerns about the quality of online education. This paper proposes a framework for comparing online learning with traditional learning using emotions, learner perception, instructors, student engagement, understanding, effectiveness, learning outcome, peer collaboration, constraints, and comparisons. Deep learning algorithms like LSTM, GRU, and RNN classify the reviews. Students are positive during online learning in higher education according to LSTM, GRU, and RNN experimental analysis. Students are becoming more comfortable with online learning environments for higher education, according to detailed survey results.

K. R. Sowmia, S. Poonkuzhali, J. Jeyalakshmi
Strategic Network Model for Real-Time Video Streaming and Interactive Applications in IoT-MANET

Gupta, Atrayee Banerjee, Bidisha Adhikary, Sriyanjana Roy, Himadri Shekhar Neogy, Sarmistha Mukherjee, Nandini Chattyopadhyay, SamiranIn this paper, we discuss how to handle the issue of energy loss, delay during processing and transmission in IoT-MANET for real-time applications such as video in surveillance networks and also interactive applications in smarthome. Here, we propose a hierarchical strategic network model to reduce the cost of computation, latency and energy consumption. The proposed model when compared with random and normal grid topology gives better performance. Results of comparison show that the proposed hierarchical model provides an average of 166.29 Mbps for bandwidth utilization, 63.92 Mbps of throughput, 0.000095 s of jitter and 12.28 % of packet loss for transferring video data in mobile conditions. Also, each supernode receives an average 6.22 frames per second out of 15 transmitted frames.

Atrayee Gupta, Bidisha Banerjee, Sriyanjana Adhikary, Himadri Shekhar Roy, Sarmistha Neogy, Nandini Mukherjee, Samiran Chattyopadhyay
Limitation for Single-equation Dependency

Shell, Michael Doe, John Doe, JaneLinear regression models are trained with standard errors. We have observed that regression models often give a low accuracy as compared to other techniques. We have evaluated the regression technique and draw a conclusion that these standard errors are the reason for the lower accuracy. The main reason of producing low accuracy is dependency on single-equation denoting relation between the dependent and independent variables. In regression model, on the basis of least square method, the affect of one variable is studied on other variable. However, there comes a time oftenly where the affect of one variable on another drastically changes due to the type of values or range of values present in the data. Such changes in the values might increase the higher accuracy of regression model. So, highly accurate regression model might not work for datasets with such variations. In this paper, we have discussed the issue of single-equation dependency to improve the accuracy. By using the ridge regression technique, we have developed a model that can produce high accuracy for multiple datasets with different variants in the values and the ranges of the values.

Michael Shell, John Doe, Jane Doe
Autocorrelation of an Econometric Model

The treatment of an econometric model requires a clearly defined sequence of tasks. The identification of the model leads us to review the literature, to justify the defined relationship between the dependent variable and the independent variables. Model estimation uses the mathematical apparatus to find the equation of fit. Once the model has been estimated, it must be properly diagnosed using statistical tests. After the diagnosis phase, one can use the model to make predictions. This contribution deals with the identification, estimation, diagnosis, and prediction phases for the treatment of econometric models. Likewise, the diagnosis is deepened by developing the problems of autocorrelation, heteroscedasticity, residual normality, multicollinearity, endogeneity, and others.

Preeti Singh, Sarvpal Singh
Trends of Artificial Intelligence in Revenue Management of Hotels

The authors provide an in-depth summary of artificial intelligence’s role in the revenue management of hotels across the globe. Focusing on artificial intelligence, the chapter draws the reader’s attention toward its usage and role in the lives of the customers and service providers in the hotel industry, which ultimately leads to the revenue management system. The chapter focuses on the current trends of artificial intelligence to generate revenue for hotels more innovatively. Some of the finest and latent features of artificial intelligence in the revenue management system have been discussed in the chapter by elucidating few examples of few renowned hotels in the world. The revenue management system of the hotel industry has been discussed briefly in the chapter. The essence of the chapter lies in the role and importance of the ever-changing and emerging trends of artificial intelligence in the hotel industry, particularly for revenue generation. Though artificial intelligence cannot replace the emotional intelligence provided by humans, it is the need of the hour to enhance and sustain in the innovative and digital era. AI rules across manufacturing and production sectors, but it also plays a strategic role in the service industry.

Smrutirekha, Priti Ranjan Sahoo, Ashtha Karki
Sentiment Analysis and Vector Embedding: A Comparative Study

Automatic intent/sentiment classification can be done with various machine learning approaches as well as methods. But the success of these techniques majorly depends on the representation of words or documents in vector space. That can be easily consumed by machine toward learning the hidden pattern of text/corpus. To achieve this, various methods have been proposed, and many are commercially accepted as well. In deep learning architecture for intent/sentiment analysis, the vector embedding plays a crucial role. It represents feature extraction. In this paper, various methods for vector embedding are discussed along with their comparison.

Shila Jawale, S. D. Sawarkar
3G Cellular Network Fault Prediction Using LSTM-Conv1D Model

Cellular network plays an important role in daily life by exploring digital world of communication. Cellular network technology continuously evolves in past decades from 1 to 5G and beyond. The evolution results in more network accessibility and data utilization. As the availability of network, mobility, and portability of cellular devices are increasing, the network traffic will also be increasing. Higher the transmission rates, higher will be the fault occurrence possibility. Monitoring network parameters and finding fault in cellular network are key factor in determining consistency of network. Cellular network which is highly dynamic than usual networks needs intelligent way of fault handling as the human over head will be unpredictable and very high. Modeling intelligent network fault identification system can simplify human efforts and improve efficiency with better accuracy. The research is on real-time data of 3G cellular network including various network parameters like uplink threshold and identifies the behavior of data usual or unusual to predict the fault occurrence. The study is on various LSTM techniques such as bidirectional LSTM, vanilla LSTM, and stacked LSTM combined with time distributed Conv1D.

N. Geethu, M. Rajesh
Deep Learning for Part of Speech (PoS) Tagging: Konkani

This is the first time that an experiment using deep learning has been attempted with a Konkani language data set. For this study, over 100,000 PoS tagged Konkani sentences were used. The f-scores for deep learning are 90.73% for training data and 71.43% for test data. These results are better than the ones previously reported for Konkani. We have provided a list of references of PoS tagging for Indian languages specifically using deep learning to place our research in perspective.

Annie Rajan, Ambuja Salgaonkar, Arshad Shaikh
DAMBNFT: Document Authentication Model through Blockchain and Non-fungible Tokens

Hard copies of documents can easily be forged, resulting in the decrease of credibility of issuing institutions and unfair use of forged documents by certain individuals as well. The model that we propose aims to authenticate documents through the use of blockchain technology, non-fungible tokens, and interplanetary file system. When a document is stored on the blockchain, a non-fungible token is created, which contains the unique address of the issuing institution and the hash of the document itself. The ownership of this token is then transferred to the document holder by the corresponding issuing authority. In this way, when someone wants to verify the authenticity of the document, they can use the address mentioned in the token to trace back the creator. If the document’s hash differs from the one stored in the token, we know that the document has been altered. Even when unauthorized users are successful in adding the forged documents to the blockchain, they will not have the same unique signature as that of the authorized institution. The proposed model allows autonomous authentication of documents using public blockchain technology.

Uday Khokhariya, Kaushal Shah, Nidhay Pancholi, Shambhavi Kumar
A Novel Hybrid Translator for Gujarati to Interlingual English MTS for Personage Idioms

Gujarat is one of the states in the western part of India, and the Gujarati language is the official language of Gujarat. The Gujarati language is more than 700 years old and is spoken by more than 55 million people around the world. A machine translation system (MTS) is needed for the communication between people knowing different languages. Idioms are used in almost all the languages. An idiom is a phrase or an expression whose meaning does not necessarily relate to the literal meaning of its individual words. The idiom generally means something different than what is directly conveyed by its individual words. Translation of idioms for any language, as with Gujarati idioms, into any other language is a challenging task. All existing MTS including Google Translate and Microsoft Bing fail to translate Gujarati idioms suitably. In the current paper, a particular category of Gujarati idioms, called the personage idioms, has been treated for detection from the input text and translation into the English language. We have deployed a dictionary-based algorithm and context-based search, respectively, for idioms having one and multiple meanings. This is a first of its kind work in the world. From a broad technical view point, this is an application of Gujarati to interlingual translation for the MTS sub-domain of natural language processing (NLP). The interlingual language is considered to be English for the present research work.

Jatin C. Modh, Jatinderkumar R. Saini
Alleviation of Voltage Quality-Related Issues: A Case Study of Bahir Dar Textile Share Distribution System

This paper presents a unified power quality conditioner (UPQC)-based compensation approach for voltage quality improvement in Bahir Dar textile share company distribution system. The existing system was modeled and simulated to evaluate its performance. The base case waveform shows that power quality is deteriorating in the factory. The voltage sag and swell at the base case are 65% and 140%, respectively, with voltage unbalance and harmonic content for a duration of 0.1 s. After using the proposed fuzzy logic controller (FLC)-based UPQC, both voltage sag and swell are alleviated by 100% and the waveform is maintained and kept at 1.0 p.u.

Dessalegn Bitew Aeggegn, Yalew Werkie Gebru, Takele Ferede Agajie, Ayodeji Olalekan Salau, Adedeji Tomide Akindadelo
Customer Perception, Expectation, and Experience Toward Services Provided by Public Sector Mutual Funds

In recent years, mutual funds have gained popularity as a means of ensuring the financial security. Mutual funds have benefited families in capitalizing on India's wealth while also contributing to the country's economic record. As knowledge and understanding of mutual funds grow, most people are reaping the benefits of investing in them. This study is focused on the various customer perceptions, expectations, and experiences toward services provided by public sector mutual funds. This study mainly analyzed the main features and major factors affecting the public sector mutual fund’s customers’ experience, expectation, and their perception. A quantitative study was undertaken through a structured questionnaire to analyze major factors that attracted customers toward public sector mutual funds, and based on the analysis, it was found that customers are satisfied because they have got better responsiveness from public sector mutual funds to customer complaints. As part of the study, many customers’ opinions and reviews have been collected. This project will help public sector mutual funds to develop new strategies to create a more customer-friendly approach for increasing the sales and improving the customer satisfaction and their experience with public sector mutual funds.

R. Malavika, A. Suresh
DigiDrive: Making Driving School Management Effective

The Indian driving school system consists of schools ranging from small-scale setups to huge franchises with branches spread nationwide. Irrespective of the scale of the school, it needs to have a management system to handle its student and instructors data to ensure seamless communication. In today’s world, technology has transformed how industries handle their data management processes. However, from this study, it was observed that the Indian driving schools follow conventional and manual methods for managing their data which generate possibilities for errors. Furthermore, it also leads to several communication problems among the users. This study explores the data management methods, and the problems faced in driving schools through contextual inquiry, semi-structured interviews, and competitive analysis. Based on the problems identified, design intervention is proposed consisting of an in-vehicle infotainment system linked to a mobile and Web application to provide a seamless management experience.

Manasi Khanvilkar, Atreya Rastradhipati, Wricha Mishra
Revamping an E-Application for User Experience: A Case Study of eSanjeevaniOPD App

Menon, Remya Vivek Rejikumar, G.E-governance initiatives are likely to succeed only if the applications created for those purposes offer an excellent user experience (UX). For UX to improve, many components associated with the application should meet expectations and make user interactions easier, intuitive, and relaxing. This study aimed to verify the user experience developed by the ‘eSanjeevaniOPD’ app and suggest ways to improve it. The sequential incident technique (SIT) revealed user concerns in every stage of the user journey and, therefore, improvement opportunities. Accordingly, a few attributes were chosen to improve UX by providing them in the best possible manner. The Taguchi experiment with ten selected attributes by capturing user perceptions identified an optimum combination of these attributes for maximum UX.

Remya Vivek Menon, G. Rejikumar
Secured Data Transmission in Low Power WSN with LoRA

Power consumption factor is very important in WSN and other IOT setup. As the nodes remain in distant locations they need to run powered by battery. When we try to send private data through it we can encrypt it. But a single key for a long period may be less secure. Hence, we need to change the key. But in a low power low processing power setup the conventional key generation algorithm may not be quite feasible as it may need a lot of processing and power. Therefore, we study low power alternative for dynamic key generation and implement it in a LoRa-based WSN.

Arabinda Rath, S. Q. Baig, Bisakha Biswal, Gayatri Devi
Safeguarding Cloud Services Sustainability by Dynamic Virtual Machine Migration with Re-allocation Oriented Algorithmic Approach

Data centres are networking platforms which exhibit virtual machine workload execution in a dynamic manner. As the users’ requests are of enormous magnitude, it manifests as overloaded physical machines resulting in quality of service degradation and SLA violations. This challenge can be negotiated by exercising a better virtual machine allocation by dint of re-allocating a subset of active virtual machines at a suitable destined server by virtual machine migration. It is exhibited as improved resource utilization with enhanced energy efficiency along with addressing the challenge of impending server overloading resulting in downgraded services. The aforesaid twin factors of enhanced energy consumption and enhanced resource utilization can be suitably addressed by combining them together as a single objective function by utilizing cost function based best-fit decreasing heuristic. It enhances the potentials for aggressively migrating large capacity applications like image processing, speech recognition, and decision support systems. It facilitates a seamless and transparent live virtual machine migration from one physical server to another along with taking care of cloud environment resources. The identification of most appropriate migration target host is executed by applying modified version of best-fit decreasing algorithm with respect to virtual machine dynamic migration scheduling model. By executing the selection algorithm, the hotspot hosts in cloud platform are segregated. Subsequently, virtual machine-related resource loads are identified in descending order with respect to hotspots. The resource loads pertaining to non-hotspot hosts are identified in ascending order. Next, the traversing manoeuvring in non-hotspot hosts queue is exercised for identification of the most appropriate host to be reckoned as migration target host.

Saumitra Vatsal, Shalini Agarwal
Survey on Cloud Auditing by Using Integrity Checking Algorithm and Key Validation Mechanism

Cloud computing can be used to access and storing data and delivery of different services over the internet. Using cloud storage, users can remotely store their data. Cloud service provider (CSP) provides data owners to store and access their valuable data in the cloud server and offers them to make use of on-demand data access without maintaining a local copy of their data. Even though this service avoids the data owners from making use of their third-party auditor, it certainly possesses serious security threats in maintaining the data owners cloud data. In addition, integrity is also an important issue in maintaining the data owners data stored in the cloud server. This survey presents an overview of integrity check and continuous auditing. The review work based on creating secure clouds by continuous auditing and cloud certification system (CCS) is used for high level security. This survey helps to provide the security by continuous auditing and overcome the integrity issues and to avoid the attacks by integrity checking algorithm and key validation mechanism. In this survey paper, various researchers’ ideas based on integrity checking and key schemes have been analyzed as literature review.

M. Mageshwari, R. Naresh
Privacy-Preserving Min–Max Value Retrieval Using Data Obfuscation Algorithm in Distributed Environment

A comprehensive and efficient p-based algorithm is being proposed to retrieve minimum and maximum of particular data field of database tuple distributed as a horizontal partitioning of single database table among the multiple servers without losing the privacy of individual server. Data obfuscation technique is used in the proposed algorithm to preserve the privacy of individual server. This algorithm is useful in many practical scenarios where privacy-preserving data computing is desirable.

Bhople Yogesh Jagannath
Intelligent Virtual Research Environment for Natural Language Processing (IvrE-NLP)

In the 21st century, natural language processing (NLP) has obtained much prominence for human–machine interaction (HMI). With this interest in natural language processing (NLP) has grown significantly, numerous NLP tools (e.g., morphology, the tagger, and a parser, etc.) have been developed all over the world. Despite having huge importance and requirements, we have noticed gaps for having a comprehensive single framework or platform, which encompass all NLP-related tools and technologies for promoting the research in NLP and sharing the knowledge and resources among NLP researchers required for understanding and building the solution for HMI. Our objective is to apply Software engineering in natural language processing with the concept of an object-oriented model by using a collection of reusable objects by defining the communication protocol, consisting of a set of rules that must be applied to exchange data between two NLP modules. We proposed state of art ivrE—A virtual environment for creating, modifying, executing, and analyzing various NLP solutions and technology. The proposed idea is broadly based on to define own ivrE-NLP object framework model that permits the developer to create, modify, and execute the application and analyze their outcomes by operations on visual representations of the modules. A variety of NLP-based applications (tools, modules, and plugins) already exist, they can publish into store available with environment so it can be used by research community at large. To develop complete NLP framework or platform, we require much more than just assembling or collecting these tools or modules at one place, no matter how good any tool or module is working individually. It requires not only the standards and a set of protocols, but also requires a compliant composition than a pre-defined algorithms and their implementation. In brief, we require a comprehensive open framework to bundle, manage, and integrate set of NLP tools, modules, components, applications, algorithms, and define their associated rules, comprehensive data structures, and knowledge.

Prashant Chaudhary, Pavan Kurariya, Shashi Pal Singh, Jahnavi Bodhankar, Lenali Singh, Ajai Kumar
Analyzing the Performance of Object Detection and Tracking Techniques

Objects are detected in computer perspective widely in many real-world applications. In case of video processing, detection and tracking of objects should be very proper and effective. Objects are detected and tracked by traditional methods such as background subtraction, optical flow, and frame differencing method. Convolution neural network, which is a deep learning-based approach, is recently adopted by many developers to identify the object. In this paper, methods of object detection are implemented, analyzed, compared, and discussed. Out of which a robust method has been suggested which satisfies the parameters of precision, recall, and accuracy, also the visualized parameters like object localization, classification, and forms a bounding box to the object are observed and analyzed. It is observed that convolution neural networks detect all relevant objects more accurately than traditional methods. CNN locates the identified object in a video frame using a bounding box that extracts the feature and trains the image for classification. Here, CNN is considered the most promising method for object detection and tracking and can be used in further study where complex work to be handled based on object detection like video inpainting or video restoration.

Suchita A. Chavan, Nandini M. Chaudhari, Rakesh J. Ramteke
A Novel Area Efficient Multiplier

Ali, Asfak Sarkar, Ram Das, Debesh KumarGalois field multiplication has received a lot of attention from researchers due to its use in encryption, channel coding, and digital signal processing. This paper proposes an area-efficient Galois field multiplier. A sequential approach is adopted here to implement the multiplier in field programmable gate arrays (FPGA). The proposed design gets a $$64.58 \%$$ 64.58 % improvement in an area with respect to 16-bit combinational implementation and $$48 \% $$ 48 % with respect to 16-bit sequential implementation. The proposed method also implemented using fixed polynomial, where the design gets a $$73.9 \%$$ 73.9 % improvement in an area with respect to 16-bit combinational implementation and $$55 \% $$ 55 % with respect to 16-bit sequential implementation. In comparison to early methods, the time delay is also decreased in most cases.

Asfak Ali, Ram Sarkar, Debesh Kumar Das
Virtual Testbed of Vehicular Network for Collision Simulation and Detection on SUMO and OMNeT++ Simulators

Vehicular safety technologies play a vital role in preventing or minimizing the impact of vehicle collisions to reduce life-threatening injuries and keep down vehicle collision-related casualties. One such application is connected vehicles, powered by vehicle-to-infrastructure (V2I) technology to enhance safety on road. It enables all the vehicles on a road within its range to communicate their speed, position, and heading direction to roadside unit (RSU) through cooperative awareness messages (CAM). This process needs three major operations. The first one is receiving the data from the vehicles, the second one detects the collision, and the third one communicates it with the vehicle in case of an impending collision. In this study, we developed a sophisticated algorithm to detect collisions. On detection of the impending collision, RSU sends a warning message to the concerned vehicle. This alerts the driver to take control measures like brake and speed limiting. Here, we implemented the intersection and rear-end collision scenarios using simulation of urban mobility (SUMO) traffic simulator and developed vehicular network (VANET) on network simulator OMNET++ . Veins framework combines both traffic and network simulator. Now using this computerized testbed, we can simulate the collision scenarios on the connected network and evaluate the timeline and data delivery rate with which the latter received the signal in order to take control actions like brake or halt the vehicle.

Tatini Shanmuki Krishna Ratnam, C. Lakshmikanthan
Artificial Intelligence-Based IoT Clustered Smart Vending Machine

Vending machines are a vital part of normal people’s everyday life in countries like Japan or the USA. India is lacking in the use of effective methodologies in vending machine space. The smart vending machine is designed to work ambiently with IoT hardware and the proprietary design architecture of the physical machine. This vending machine is designed to reduce the mechanical complications present in the current machinery and use technically advanced systems based on IoT; which will help in remotely interfacing with the machine as well as intelligently compiling the data which our artificial intelligence system interprets, giving us predictions for customers as well as the vendor. The machine will be connected to the cloud, responsible for comprehensive data collection and processing. Our system will be having a cluster of vending machines that will be interconnected to each other. The singular vendor can view the data generated by all these machines on a single console and will be able to control and monitor the aspects such as product management, machine state check, and many more from the console itself. The machine will not be having any physical interface for users, and we have designed an utterly end-to-end architecture for this system which will revolutionize the vending machine space in the industry.

Ravivanshikumar Sangpal, Suhas Khot, Pratiksha Vallapure, Rajkumar Mali, Rasika Kumbhar, Aparna Hambarde
Artificial Intelligence-based Vehicle In-Cabin Occupant Detection and Classification System

Vehicle safety is the primary and necessary aspect of the automobile industry. An airbag is one of the passive safety systems available in an automobile. However, the airbag deployment needs to be controlled to avoid accidents due to it. In low-velocity crashes, the injury caused by airbag deployment is higher than the impact inside the vehicle. Children are more vulnerable to airbags when they sit near airbag housing without proper seatbelts or child seat arrangements. Deployment of the airbag when no occupant is sitting on a seat is unnecessary. So, it is important to detect the occupants’ presence in a seat and their classes such as a child or an adult. The primary aim of the paper is to detect the occupants’ presence and classify them into different classes. The occupant classes used are child and adult. In this project, we developed a technique to identify the occupants’ presence and verify the data from one sensor using another. We collected the image data of the occupants using a camera in a sedan and hatchback vehicles. We analyzed the images using a deep learning algorithm. The output classified the occupants as child and adult. A load cell sensor mounted on the seat was used to measure the weight of the passenger. This data was used to confirm the occupant classification. We evaluated the model detection and classification performances with the parameters such as precision 0.95, recall 0.97, and F1-score 0.96 for image dataset, and we got 0.73 as a classification accuracy for load cell dataset. Finally, we compared both the model performances.

M. Tamizharasan, C. Lakshmikanthan
A Hybrid Approach Towards Machine Translation System for English–Hindi and Vice Versa

With the rapid progress in the technology and data in the public domain, the machine translation and data science have made remarkable progress. In this paper, we discuss our specific use case of developing machine translation system for English to Hindi and Hindi to English language translation. For this system, we have used the daily proceedings of the Lok Sabha as data and developed NMT-based machine translation system on the top of already available rule-based machine translation system. Developed system has been evaluated using bilingual evaluation understudy (BLEU) as well as the human evaluation metrics using comprehensibility and fluency. In machine translation (MT), there is the trend of measuring post-editing time, and thus, we have also evaluated our system by measuring post-editing time using open-source tool.

Shraddha Amit Kalele, Shashi Pal Singh, Prashant Chaudhary, Lenali Singh, Ajai Kumar, Pulkit Joshi
Gender Issues and New Governance: Managing Menstrual Health and Hygiene with Emerging Multimedia Options for a Progressive Economy

A nation-building process needs to be comprehensive and progressive with strong backward and forward linkages. Hence, insight into the evolution and foresight driven endeavors to change, grow and develop further become the cornerstone for sustainable development. Women-centric issues, especially those related to health, hygiene and sanitation will play a cataclysmic role in this broad context. Here the design and implementation of an effective outreach communication for social mobilization based on substantial ‘reversal of learning’ is expected to lead toward more responsive, potent and sustainable governance. We are witnessing a progressive global interest in the domain and much of so in India. What is still missing is an overarching, orchestrated mechanism, framework and governance which will curate novel initiatives and ideas, filter redundancy and repetition and harness emerging multimedia options by tapping the state-of-the- art cognitive and immersive technologies to loop the population into a dialogic communication for inclusive growth. This innovative, customized, mediate communication paradigm will disseminate knowledge to spread awareness for better menstrual management practices across the nation to promote a healthy and efficient workforce to support a healthy economy.

Suparna Dutta, Ankita Das
An IoT-Based Smart Parking System for Smart Cities

There has been an exponential growth in the vehicle population in recent times, adding to the metropolitan traffic. This increase in the number of vehicles has led to the problem of inadequate parking spaces resulting in traffic congestion. To address this challenge, a smart parking system is proposed in this paper, which makes use of TIME RESOURCE SHARING to effectively utilize the parking spaces based on peak demand time and enables prior identification and reservation of parking space with the help of unique identification. The system periodically updates the parking status. The interaction of the driver and the owner of the parking space takes place through the application connected through the cloud. The related work in this area has also been referred. The proposed system has the potential to transform the current parking method and alleviate the traffic congestion caused by insufficient parking space.

Venkatesh Mane, Ashwin R. Kubasadgoudar, Raghavendra Shet, Nalini C. Iyer
Agriculture Stakeholder’s Information Need—The Survey in Gujarat, India

Agriculture plays an important role in Indian economy. In recent era applications and technology usage is drastically increased. To fulfill the demand in Agriculture, it is important to know the information need of the farmer. Farmer’s Information Need can be satisfied by the varied human stakeholders in Agriculture or by the technology use. We have prepared Agriculture Questionnaire to know the farmer’s information need. Our Agriculture Survey is filed as copyright, detailed in [1], We have conducted survey to the varied demography of the farmer on different time and at different location. To improve the accuracy of the survey, we have interviewed the stakeholders on one-to-one basis directly. Our survey to the agriculture stakeholders of Gujarat also enables and facilitate us examining the channels of information communication, gathering information need and knowing varied need in natural Guajarati language. Facts and figures of the survey leads to the necessity of intelligent retrieval system.

Axita Shah, Jyoti Pareek
System Design and Implementation of Assistive Device for Hearing Impaired People

This paper discusses the design, implementation and development of assistive device for hearing impaired. Sign language is one of the oldest and most natural form of language for communication, but since most people does not know sign language. Finding interpreters for different group of people is difficult. To treat the deaf people as one among the society, a user-friendly assistive device is necessary. This paper provides the details about the implementation of standalone interpreter using transfer learning for finger spelling-based American Sign Language using Raspberry Pi. A Graphical User Interface (GUI) is created and tested for establishing a two-way communication to convert text into sign language. A language barrier is created using an assign language structure that is different from normal text. Hence, the users will now depend on the vision-based communication that will be able to bring normal people, deaf and mute people on the same grounds of interaction. There is a need to build a possible sign language translator, which can take communication in sign language and translate them into written and oral language. Such a translator that uses neural networks for finger spelling-based American Sign Language would greatly lower the barrier for many deaf and mute individuals to be able to better communicate with others in day-to-day interactions. In proposed method, hand gesture is first passed through a filter, and after the filter has applied, the gesture is passed through a classifier which predicts the class of the hand gestures. This method provides 95.7% accuracy for the 26 alphabets and 0, 1, 2 numerical.

P. Nikita, B. H. Shraddha, Venkatesh Mane, N. Shashidhar, P. Vishal, Nalini C. Iyer
Internet of Things Security: A Blockchain Perspective

Since the arrival of computers in 1950s, technology has leaped over many generations at a rapid pace. Computers have found their use in every human activity, namely scientific experiments, industries, transportation, medicine, healthcare, home management, etc. Internet of things (IoT) is one such computing paradigm that aims to assist and automate various trivial and non-trivial aspects of these activities. IoT generates a huge amount of data at a fast rate. The data can be trivial, personal, or sensitive to someone or an organization. Hence, securing IoT data and its aggregation is of utmost importance. The paper gives a brief introduction of IoT in terms of its security aspects to understand the various facets of malicious attacks possible on it. Then, how blockchain, an ever-popular security suite, can be used to provide security to IoT in its various integrations is presented. Finally, some future research issues are discussed towards blockchain-IoT future.

Mohammad Luqman, Arman Rasool Faridi
Fuzzy C-Means Clustering of Network for Multi Mobile Agent Itinerary Planning

Mobile agent (MA) works potentially efficiently in reducing network bandwidth consumption for distributed computing. In an MA-based system, a small-sized processing code is transmitted through the network rather than the raw data. Using strong migration capability, MA’s processing code is being executed at each targeted node. Subsequently, useful and significant information is delivered to the intended user. Despite its benefit, MA has its issues also. Finding the appropriate number of MAs to be dispatched, the set of targeted nodes and their sequence of migration are the major issues related to multi-mobile agent itinerary planning (MIP). This paper tries to find a suitable number of MAs by considering the size of the data payload of a single MA and the load to be carried out by the MAs from the whole network. Moreover, this paper gives an idea of partitioning the given network into load balanced and nonoverlapping clusters using a fuzzy c-means clustering algorithm.

Nidhi, Shuchita Upadhyaya
Acceptance of Eco-Friendly Substitute of White Skim Coat (Wall Putty) in Mumbai

Wall putty available in the market is white in colour and made from perishable raw materials. It is possible to have a substitute product, black in colour and made from using waste powders from quarry and construction. Replacement of the current white putty with this substitute black putty is technically possible. Our research is on the acceptance of this product. On conducting a survey in Mumbai and surrounding suburbs, we asked questions on the awareness, acceptance of self and perceived acceptance of the industry. The results are compiled and analyze. The resistance shown by the various influencers of the industry is significantly on the higher side. However, as researchers we feel that there is certainly a possibility to make this change and develop the market in the direction of usage of black putty which is actually an eco-friendly product using waste materials. Another important benefit of using this product is that it actually reduces pollution. The waste product used as raw material is very fine, and these particles create air pollution. We have also studied and worked out the cost of this substitute product which is actually cheaper than the current white putty. The research has further scope to be carried out in other areas of the country and interacting with more population.

Mandar Anil Chitre, Shivoham Singh, Manuj Joshi
Economics of Immutability Preserving Streaming Healthcare Data Storage Using Aggregation in Blockchain Technology

The contemporary technology landscape is big data oriented and driven by analytics. The paradigm shift fueled by rapid data generation and higher storage capabilities has resulted in a massive transition across data storage technologies. Traditional database management systems being used in record-based transactional applications are paving the way for new rapid engagement-based data stores leading the newer applications. This transition is a result of a sharp growth trajectory witnessed across many parallel technology landscapes including cloud for storage and processing, Internet of things (IoT) for rapid data generation and transmission, support for unstructured, semi-structured and structured data by social media platforms, and polyglot persistence supported by NoSQL, to name a few. The emerging applications thrive on insights acquired by data analytics, and immutability of data storage is expected to be one of the key factors in growth of forward leaning enterprises. There has been a constant rise in application development requiring a full history of transactions, which is both trustworthy and traceable. Such applications mandate the append-only nature of the data storage to support analytics and trust. Immutability is also the underlying premise of popular cloud native storage systems. This paper explores blockchain technology as a solution for privacy and disclosure compliance in stored healthcare data. Blockchain as a data structure is primed to store only small amounts of data to maintain the properties of immutability, tamper proofing, security, and transparency in applications and is not suitable for storing big data. However, healthcare blockchain implementations have several possibilities for storage management, and the paper presents a comparative analysis of the aggregation-based storage with IoT-based streaming e-healthcare application data for the use case. The in-depth analysis for the potential blockchain storage in the paper includes the cost factors, in addition to immutability and privacy preservation, and the results show that costs may be saved up to an order of magnitude 300 with aggregation.

Sachin Gupta, Babita Yadav
A Comparative Cost Analysis of Organizational Network Security Test Lab Setup on Cloud Versus Dedicated Virtual Machine

The global network infrastructure spectrum is witnessing its fastest growth since the last decade with concurrent rise in cloud computing, Internet of things (IoT), and edge computing. There has been a multitude of heterogeneous networking devices spanning different configurations and using a variety of access methods. A parallel evolution of the network infrastructure security is happening with increasing attempts to exploit the security vulnerabilities in mission critical cyber-assets of organizations. Several organizations invest heavily in security research using lengthy and cryptic mathematical models while ignoring the practical network implementation situation and focus only on the monetary implications of the attack and defense. Attack tree has evolved as a convenient and cost effective way of plotting the network in which an attack may take place and can also help organizations understand the way it can be defended. Attack trees combined with the MITRE ATT&CK framework are widely used for crown jewels risk assessment globally. However, the major challenge for information security experts using the attack tree methodology lies in manually creating the attack tree and plotting all the crown jewels and perimeter network so that it can be defended from attackers. We propose a test lab setup for simulation and attack tree generation, which can be used in conjunction with the MITRE ATT&CK framework and allow us to create and assess various attack scenarios while providing flexibility in subnet configuration and movement, addition or removal of networking devices. The lab can be cloud hosted with a popular cloud hosting on Microsoft Azure or may be created on a VM within a dedicated high-resource machine to be used as a portable testbed. The results indicate that both services have their own pros and cons based on the hours of usage, and the dedicated resource VM testbed may perform better in a low-risk potential small network while the cloud-based approach is useful for the scalable organizations with high-threat potential.

Sachin Gupta, Bhoomi Gupta, Atul Rana
Insights into the Black Box Machine Learning Models Through Explainability and Interpretability

Artificial intelligence (AI) and machine learning (ML) technologies are considered to be the Holy Grail for the researchers across the world. The applications of AI and ML are proving disruptive across the global technological spectrum, and there is practically no area which has been left untouched by these technologies right from computer science to manufacturing, healthcare, insurance, credit ratings, cybersecurity, and many more. It would not be an exaggeration to say that it is the next big thing after the advent of the Internet and potentially holds a similar impact in touching the lives of human beings. Whilst most researchers using machine learning in research across diverse domains do not need to look beyond the model abstraction for their work, the need for understanding what is happening beneath the surface is sometimes necessary. This becomes especially important in the cases where the predictions are too good to be apparently true, and the researcher running the model is not sure about its validity as the logic for prediction is obscure. The process of feature engineering brings in more accuracy to predictions, but in the absence of intuitive background information regarding the features, the task gets more challenging. The scientific reasoning has been driven by logic through ages, and the scientist community remains sceptical of the results unless they can extract useful insights from the black box ML models. The paper applies five popular explainability algorithms being used by the research community to demystify the abstract nature of ML black box models and compare the relative clarity of the insights being provided individually by each from a practitioner’s perspective using the publicly available UCI wine quality dataset.

Sachin Gupta, Bhoomi Gupta
An Improved AODV Routing Algorithm for Detection of Wormhole and Sybil Attacks in MANET

Wireless sensor network is a promising technology in the current scenario due to its wider area of research. With the advancement of technology, communication in mobile ad hoc networks has become easier and efficient with some vulnerability. Secured system communication is required, but with the evolution and advancements in the technology, threats also increase. During communication, various kinds of attacks may occur in the mobile ad hoc network. In this paper, wormhole attack in AODV protocol is discussed. Different parameters affect the working of protocols; here parameters end-to-end delay and PDR ratio are discussed with the main aim to detect the wormhole attack and provide an efficient solution to minimize the risk of attack.

Shreeya Mishra, Umesh Kumar, Komal Mehta Bhagat
An Ensemble Model (Simple Average) for Malaria Cases in North India

Malaria is an infectious disease borne due to mosquitoes that attacks humans and other animals’ bodies. Malaria is a part of the plasmodium group caused by single-celled microorganisms. This study proposes the use of ensemble model using the three regression algorithms that are linear regression, support vector machine (SVM), and auto-Arima techniques and comparing their results. Predictions of plasmodium virus cases are made with the use of linear regression, support vector machine, and auto-Arima algorithms. The accuracy of prediction is measured by calculating the explained variance score, mean squared error rate, and root mean squared error rate. Our aim is to get better prediction results compared to the individual algorithms by combining the results of these individual models. The proposed work determines the accuracy of linear regression, support vector machine, and auto-Arima and ensembles together to find the trend of prediction using simple Average. A comparison of performance among the three regression techniques indicated the SVM model performs the best and has small RMSE and MAE values. But, by introducing the technique of ensemble modeling using simple average, combining the prediction of these three algorithms results in the lowest RMSE and MAE values.

Kumar Shashvat, Arshpreet Kaur, Ranjan, Vartika
IoT-Based Smart City Management Using IAS

One of the most important factors in modern information systems of the Indian administrative agencies is the lack of easy access to information through organizations outside of the organization, with the name of the machine-guns in the app. In spite of the abundance of data, and they are locked into proprietary formats, or may not be available due to the lack of an API. In this report, we will focus on the architectural designs that make it easy to share data. Therefore, we focus on a data-based information, see the reference architecture, rather than focusing on the application and communication. Internet of Things (IoT) is a new technology that has emerged in the last couple of years, and it promises to be a continued explosion of data sources. With this technology, it will, in essence, a single “thing” to begin with, the production and transmission of data. For example, each and every lamp post, street lamp, trash bin, power transformer, etc., almost anything that a person can to be created in order to be constantly informed about their disease, their perceptions, and their patterns of use, etc. These data, which, in turn, can be used to lower the level of the business processes of the administration, in order to ensure fast and efficient value-added services. In this report, we will limit ourselves to the architectural design for the integration of new IoT-based systems in a Smart City infrastructure.

Pranali Nitnaware, Supriya Sawwashere, Shrikant V. Sonekar, M. M. Baig, Snehal Tembhurne
Big Data Analytics and Machine Learning Approach for Smart Agriculture System Using Edge Computing

With the development of the Internet of Things (IoT) and machine learning technology, a smart agriculture environment produces more agricultural land and crop-associated data for knowledge discovery systems. Machine learning decision-making algorithm is applied to discover hidden knowledge patterns from the agricultural data stored in the distributed database. Big data analytics extract useful information from the large, distributed, and complex datasets, which helps the farmer to increase crop yield and quality of the production. The edge computing node collects crop data and land environment data from the agricultural lands using a different kind of IoT sensors. The predicted smart agricultural knowledge pattern can provide needed information to the farmers and other users like an agent, agriculture officers, researchers, and producers to get more profit. Cloud and fog computing provides efficient distributed data storage for big data and execute dynamic operations to predict business intelligence facts to increase production and minimize natural resource utilization. We have compared traditional data mining techniques with the business analytical tool hybrid association rule-based decision tree (HDAT) MapReduce approach for implementing decision tree algorithm to predict and forecast the future needs of the farmer to increase the profit and reduce the resource wastage.

U. Sakthi, K. Thangaraj, T. Poongothai, M. K. Kirubakaran
Real-Time Object Detection System with Voice Feedback for the Blind People

Lots of people suffer from temporary or permanent disabilities, and one of them is blindness or the visually impaired. According to the World Health Organization (WHO), more than 1 billion people suffer from blindness. Technology may assist blind and visually impaired persons in a number of ways; objects detecting is still a difficult process. There are so many techniques available for object detection. But, the accuracy and efficiency of detection aren’t good. Therefore, in this paper, authenticate object detection in real time using the YOLO-v3 and deep learning techniques. The main focus of this paper is to create a smartphone application that is cost-efficient with high accuracy.

Harshal Shah, Meet Amin, Krish Dadwani, Nishant Desai, Aliasgar Chatiwala
Bird Video Summarization Using Pre-trained CNN Model and Deep Learning

In the past few years, the forms of data have changed drastically from the text formats to images and today, most of the data are available in the video format. With this, there is a huge demand in the techniques that can provide the overall summary of the video. In this paper, we present the summary of birds that are identified from the large datasets using convolutional neural network (CNN). CNN is one of the best image processing and video processing model. The CNN model that we have used for the task is pre-trained AlexNet. The paper clearly proves that the work proposed recognizes the various kinds of birds from the inputted video with accuracy level ranging between 85 and 99%. At last, we provide the overall summary in terms of start time and end time of existence of each bird in the video.

Rachit Adhvaryu, Viral Parekh, Dipesh Kamdar
Generation of Seismic Fragility Curves for RC Highways Vulnerable to Earthquake-Induced Landslides Based on ICT

In steep terrains, major highways are laid along steep slopes that are vulnerable to slope failures. More often than not, such mountainous regions tend to be seismically active. Structures located in steep slopes are vulnerable to earthquake-induced landslides and affect the integrity of the highways. We analyze two major landslide events that disrupted major highways in the past and generate fragility curves to estimate the hazard in such a scenario. We derive the curves based on landslides that damaged the highway by measuring the distance from the source of rockfalls. Our results indicate that the RC within 3–5 m from the crest of a landslide tend to be at a greater risk. In an event of higher magnitude, the fragility curves estimate lower probability of damage. This could be because the shaking itself causes enough damage than the triggered landslides. It was also observed for a distance of 10 m from a slope crest in a low-magnitude earthquake, the damage is the most. This shows that the momentum of the triggered rockfall increases due to increased distance and causes substantial damage on the RC highways. Damage caused by Taiwan rockfall event which occurred in April 2019 was assessed using these curves and was found to be moderate. Fragility curves generated in this study can be used to estimate the damage in a future scenario where there are highway constructions along slopes that are prone to rockfalls.

Aadityan Sridharan, Sundararaman Gopalan
Influence of Student–Teacher Interrelationship on Academic Achievements: A Logit Model

The main concentration in the present work is given to identify the effect of the student–teacher relationship on the academic achievement of university students. For this purpose, data is collected from 292 students of various universities situated in Jaipur, India. The survey was conducted from January 2018 to April 2018. A questionnaire of 36 questions was filled by all respondents enthusiastically, out of which six questions are demographic. Responses have been collected on a five-point Likert scale. In the sample, 55.5% male and 44.5% students are included with 28.1% day-boarding and 71.9% hosteller students. Maximum 52.7% of students belong to the third year, and a minimum of 2.1% students are included from fourth year from various branches like engineering, science, arts and management. The reliability of the questionnaire was checked by Cronbach’s Alpha, and logistic regression analysis is performed. Two logistic regression models are developed concerning gender and locality of students, and odd ratios are identified. Finally, it is recommended that findings can be used to enhance the academic achievements of the students.

Monika Saini, Asha Choudhary, Ashish Kumar
Smart Water Metering Implementation

Throughout the past years, the Indian government is putting its best efforts in developing smart cities by applying smart solutions for infrastructures and services. Smart water management is one of the key service areas where now it is a high time to work for. Many industries and researchers are contributing to incorporate smart techniques for smart water system like sensor-based monitoring, real-time data transmission and controlling, leakage management, water distribution management and many more. Nevertheless, the design and proposal of such a smart water system are still not fairly standardized. The enormous applications of smart water management still do not have systematic framework to guide real-world design and deployment of a metering system. To address this challenge, a framework is designed and tested with prototype implementation. This paper mainly focuses on minimizing smart water metering application development efforts. The framework uses open-source technologies and has mechanisms like device installation, water usage statistics, bill payments and a distributed data transmission architecture.

Urja Mankad, Harshal Arolkar
Artificial Bee Colony Optimization-Based Load Balancing in Distributed Computing Systems—A Survey

Distributed computing allows the interoperability of components in distributed system in which software or hardware components located at networked computers coordinate and communicate their actions by message passing. The tasks in such a system are carried out independently. In distributed computing systems, load balancing is one of the issues, which is a means to distribute the tasks such that the computational nodes are neither overloaded nor underloaded, and the performance of the system is improved. Among the various solutions proposed for load balancing, metaheuristic-based algorithms are one of them. This paper discusses the variants and recent developments of artificial bee colony optimization algorithm for solving load balancing in distributed systems. As a result of load balancing, various performance metrics measured are throughput, response time, makespan, CPU utilization, memory utilization, and network utilization. The performance of the optimization algorithms is also measured with the time taken to converge in finding the optimal solution for load balancing.

Vidya S. Handur, Santosh L. Deshpande
Distributed and Scalable Healthcare Data Storage Using Blockchain and KNN Classification

Medical data from patients is saved electronically in the healthcare industry. The paper suggests storing medical data on a distributed on-chain utilizing InterPlanetary File Storage (IPFS) and blockchain. The use of InterPlanetary File Storage can help an organization manage the storage space efficiently. The proposed method gives details about the intersection of artificial intelligence and blockchain. This proposed approach eliminates the requirement for a trusted centralized authority, intermediates, and transaction records, resulting in increased efficiency and safety while retaining high integrity, scalability, reliability, and security. It also emphasizes smart contracts to manage and control all interactions and transactions between all healthcare participants. Using artificial intelligence, a revolutionary new system can be built in which a huge amount of medical data can be processed efficiently and can be used to predict diseases. The paper utilizes the assistance of binary classification method by K-nearest neighboring algorithm for machine learning to foresee the patient’s heart health. Using InterPlanetary File Storage (IPFS) improves the storage system framework by making data distributed and, subsequently, more reliable, scalable, and accessible to the clients.

Manjula K. Pawar, Prakashgoud Patil, Amit Singh Patel
A Sustainable Green Approach to the Virtualized Environment in Cloud Computing

Cloud technology is a dynamic industry of communication and information technology (especially the Internet) that has revolutionized current computing while also introducing new environmental concerns. The cloud is a huge improvement over the conventional computing as it has redesigned the way businesses operate. With the cloud, the information is virtualized and reduces the need of infrastructural based model. Businesses can operate through this tangible area, significantly reducing energy consumption and the need for excessive material resources. In recent years, computational technologies and ideas have transitioned to distant data centers and pay-per-use hardware and software solutions. However, as the number of information centers supporting cloud-based applications grows, vast quantities of energy are consumed, resulting to high prices and greenhouse emissions in the atmosphere. Going green emerged as a viable solution to this challenge. Green computing is a study field that intends to overcome energy and climate issues. Green sustainable cloud technology aims to accomplish not only effective processing and computing environment usage, but also to improve energy efficiency. This paper will address the possibilities of sustainable cloud technology, emerging developments, and strategies to increase data center efficiency, which also will minimize carbon emissions. This creates research hurdles when such power devices are necessary to reduce the negative environmental effect of cloud technology.

Anjani Gupta, Prashant Singh, Dhyanendra Jain, Anupam Kumar Sharma, Prashant Vats, Ved Prakash Sharma
Threats and Challenges of Artificial Intelligence in the Healthcare Industry

The authors provide a crisp yet in-depth summary of the relevance of artificial intelligence (AI) in providing healthcare solutions. Artificial intelligence is a relatively new concept in the field of health care. AI aids in the prediction of disease patients for medical procedures. Patients, pharma companies, health services, insurance companies, and medical institutions benefit from AI’s application in health care. Artificial intelligence supports different concepts, counting computing, computer program improvement, and information exchange. Machine learning, profound learning, normal dialect generation, discourse acknowledgment, robots, and biometric distinguishing proof are illustrations of artificial intelligence’s innovation. Artificial intelligence is used in a variety of areas, including health care, manufacturing, and business. It is also used in the automotive industry. The authors have discussed the current scenario of artificial intelligence and the threats and challenges posed for AI in the healthcare industry.

Priti Ranjan Sahoo, Smrutirekha, Mou Chatterjee
Exploiting Tacit Knowledge: A Review and Possible Research Directions

Saini, Pawankumar Chitrao, PradnyaIn the corporate world, tacit knowledge is becoming recognised as a differentiator. The value and necessity of focusing on leveraging tacit knowledge are highlighted via a literature study. The important themes covered in this review are the knowledge generation process, information sharing, behaviours that influence tacit knowledge sharing, sharing strategies, and so on. The research examines the different factors that lead to the industrial exploitation of tacit knowledge. To address the research gaps identified, the authors attempt to propose possibilities for future research in this area of discourse. The authors feel that the provided issues for future research will aid academics in further exploring the potential of tacit knowledge exploitation in the sectors, in addition to the research work done by many scholars in the tacit knowledge domain. The authors are conscious that this article has certain limitations because it focuses on tacit knowledge in the industry, although the impact of tacit knowledge in other sectors needs to be investigated as well. This study could lead to new approaches for businesses to tap into the potential of tacit knowledge held by their staff through knowledge production and sharing.

Pawankumar Saini, Pradnya Chitrao
An Efficient Approach to Stamp Verification

Stamps have become one of the most important security features in big companies where huge amounts of documents need processing daily. The stamp attached to a document is used to determine the authenticity of that document so that it is necessary to identify whether a stamp is forged or genuine. However, nowadays, it is easier for the general public to forge stamps. This paper presents a practical approach for stamp verification, based on the three stages process similar to some previous work: stamp segmentation, classification stamp or non-stamp, and stamp authenticity verification. In each stage, this work tries and tests new algorithms/methods to give a new way of solving the problem in each stage. Firstly, in our approach, an unsupervised learning machine method is implemented to detect all the objects in the input image, so all the regions including stamps and text are extracted. Next, two separate models of support vector machine classification are constructed. The first one is to distinguish between stamps and other objects in a document. The second model will determine the object which was classified as stamps in the first model whether it is genuine or not. The results show that this approach can perform the stamp verification tasks effectively.

Ha Long Duy, Ha Minh Nghia, Bui Trong Vinh, Phan Duy Hung
Fragmented Central Affinity Approach for Reducing Ambiguities in Dataset

It is always a known fact that the role of data and its purity is very crucial in the data mining. The key role of data in the data mining is related from decision-making. It is well-known fact that if data are impure, then result will be a false picture. This crucial stage is also known as the ambiguities in datasets. Anomalous or irregular value in database is solitary of the biggest problems faced in data analysis and in data mining applications. Data preprocessing for the data mining is a key phase which is crucial place where ambiguities of database can be reduce or remove. The present study proposed an algorithm which tries to solve the problem related to an anomalous and irregular values, i.e., outliers, inliers, and missing values from a real-world imbalanced database. The study projected is based on the fragmented central affinity approach for reducing ambiguities in dataset.

Tanvi Trivedi, Mahipal Singh Deora
Correction to: Human Activity Recognition Using LSTM with Feature Extraction Through CNN

Correction to: Chapter “Human Activity Recognition Using LSTM with Feature Extraction Through CNN” in:Y.-D. Zhang et al. (eds.), Smart Trends in Computing and Communications, LectureNotesin Networks and Systems 396,

Rosepreet Kaur Bhogal, V. Devendran
Smart Trends in Computing and Communications
Yu-Dong Zhang
Tomonobu Senjyu
Chakchai So-In
Amit Joshi
Copyright Year
Springer Nature Singapore
Electronic ISBN
Print ISBN

Premium Partner