Manish R Joshi
@nmu.ac.in
Professor, School of Computer Sciences
Kavayitri Bahinabai Chaudhari North Maharashtra University
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
Snehalata Bhikanrao Shirude and Manish Ratnakar Joshi
Emerald
Purpose Free Open Source Softwares (FOSS) witnessed the development of many very good alternatives to proprietary softwares. These free softwares can be localized in several local languages. This paper aims to illustrate a very interesting empirical investigation on FOSS. Several significant benefits of localization are described in introduction and subsequent sections. Design/methodology/approach Although the localization process is standard and well documented for most of the FOSS, it is a more complex task as it involves coordination among developers, linguists and domain experts. Hence, a very few open source softwares are successfully localized in Indian languages. In this paper, the authors present an approach that they have used for GIMP (GNU Image Manipulation Program) software Marathikaran (localization in Marathi language) project of by Rajya Marathi Vikas Sanstha of Maharashtra Government (RMVS), India. Findings This localization project has been described by RMVS as a pilot project that would guide such similar localizations in many other Indian languages for other popular open source softwares. Social implications The localization work overcomes the general misconception that regional languages are good only for communication (Boli Bhasha) but cannot be used for dissemination of knowledge (Gyan Bhasha). This work is notably contributing to language preservation, language revitalization and Digital India Initiative. Originality/value This work is the pioneering work in this domain for Marathi language with respect to GIMP. The authors presented systematic steps used to localize the GIMP software in Marathi language (from 2% to 100%).
Manish Joshi and Sangeeta Chakrabarty
The Royal Society
Dance is an art and when technology meets this kind of art, it is a novel attempt in itself. Many researchers have attempted to automate several aspects of dance, right from dance notation to choreography; from dance capturing to dance generation. We define and illustrate the concept of ‘Dance Automation’ in this paper. Furthermore, we have encountered several applications of dance automation like e-learning, heritage preservation, medical therapy, etc. Despite decades of continuous attempts by many researchers in various styles of dance all round the world, we found a review paper that portrays the research status in this area of ‘dance and computers’ dating to 1990 (Leonardo 1990 Computers and dance: A bibliography , pp. 87–90). Hence, we decided to compose a comprehensive review article that showcases several aspects of dance automation and document contributions of researchers in marrying creativity with logic. This paper is an attempt to review research work reported in the literature, categorize and group significant research work completed in a span of 1967–2020 in the field of automating dance. We have explicitly identified six major categories corresponding to the use of computers in dance automation, namely, dance representation, dance capturing, dance semantics, dance generation, dance processing approaches and applications of dance automation systems. We classified several research papers under these categories according to their research approach and functionality. With the help of proposed categories and subcategories, one can easily determine the state of research and the new avenues left for exploration in the field of dance automation.
Theyazn H. H. Aldhyani, Manish R. Joshi, Shahab A. AlMaaytah, Ahmed Abdullah Alqarni, and Nizar Alsharif
Hindawi Limited
According to the World Health Organisation, three to five million individuals are infected by influenza, and around 250,000 to 500,000 people die of this infectious disease worldwide. Influenza epidemics pose a serious public health threat. Moreover, graver dangers are encountered with influenza subtypes against which there is little or no preexisting human immunity. Such subtypes of influenza have the potential to cause devastating epidemics. Thus, enhancing surveillance systems for the purpose of detecting influenza epidemics in an early stage can quicken response times and save millions of lives. This paper presents three adapting intelligence models: support vector machine regression (SVMR), artificial neural network using particle swarm optimisation (ANNPSO), and our intelligent time series (INTS) to predict influenza epidemics. The novelty of the current study is that it proposes a new intelligent model to predict influenza outbreaks. The INTS model combines clustering with a time series model to enhance the prediction of influenza outbreaks. The innovation of our proposed model integrates the results obtained from the existing weighted exponential smoothing model with centroids obtained from clustering. We developed a surveillance system for influenza epidemics using Google search queries. The current research is based on a weighted version of the Center for Disease Control and Prevention influenza-like illness activity level obtained from the Center for Disease Control and Prevention data, as well as query data obtained from the Goggle search engine in the USA. The influenza-like illness data was collected from January 4, 2009 (week 1), to December 27, 2015 (week 52), stretching across a total time span of 312 weeks. Google Correlate was used to select search queries related to influenza epidemics. In total, 100 search queries were obtained from Google Correlate, 10 of which were better and more relevant search queries selected in this study. The model was evaluated using online Google search queries collected from Google Correlate. Standard measure performance MSE, RMSE, and MAE were employed to estimate the results of the proposed model. The empirical results of the INTS model showed MSE = 0.003, RMSE = 0.036, and MAE = 0.0185, indicating that the errors of the proposed model are very limited. A comparative model of predicting results between the INTS model, alternative Google Flu Trend (GFT), and autoregression with Google search data is also presented. The proposed model outperformed the existing models.
Yogita Patil and Manish Joshi
IEEE
Trend prediction of the volatile stock market has been an interesting and challenging task for many researchers over many years. In this paper, we present how rough set-based BIRCH clustering can be used to develop stock data prediction model. The proposed model augments clustering with a popular technical analysis method called candlestick. BIRCH clustering algorithm is used to group stocks of varied sectors by taking into consideration the previous few days volatility. Further cluster analysis is carried out to predict stocks movement for next trading day. The proposed prediction model is different from existing models as it works on all NSE stocks from varied sector. Our model outperforms models that merely using clustering or candlestick techniques.
Manish Joshi
IEEE
Personalization in e-services is desirable and large numbers of professional players are ensuring that personalization must be included as a web service for users. Moreover, recommender systems can perform effectively only with the support of personalization. Personalization has gained momentum in the service sector including education. With the advancement of the concept of ‘Teaching with Technology’, industry is inching forward to provide personalized learning contents on e-learning platforms. Personalization can be offered to a learner by ana-lyzing learning behavior, cognitive skills, learning style etc. of a learner. Most of the researchers have attained personalization especially in distance mode of e-learning using Adaptive Educational Hypermedia Systems (AEHS). Different aspects of personalization that demonstrate a paradigm shift from synchronous to adaptive approach of e-learning are being explored and experimented by many researchers. In this paper, we present our experiments of delivering learning objects (LOs) to learners that suits to the learners learning style. A personalized instruction delivery mechanism is developed that matches Learning style of a LO and the learning style of a learner. We demonstrate how such matching is ensured. We present the design of the Intelligent Tutoring System that we have developed and further discussed the learning style driven content delivery personalization.
Yogita S. Patil and Manish R. Joshi
Springer Singapore
Varsha M. Pathak and Manish R. Joshi
Springer Singapore
Manish Joshi, Bramah Hazela, and Vineet Singh
IEEE
Internet of Things, A whole new concept in the world connecting through web now focusses on connecting real objects with the internet and providing a whole new dimension of human evolution that enables us to make decisions more precisely, smart actions using smart objects on single tap. This way of connecting people's life will give us the whole new bunch of applications. This paper focusses on providing collaborative approach of using data mining techniques with the Internet of Things where decision providing system based on smart decisions and actions are actually backed up by knowledge extraction process. This paper uses Hungarian database under heart disease dataset as a case study, where different data mining algorithms like decision tree, k means clustering and naïve bayes are applied generating eye opening facts about the datasets with the help of open source data science platform Rapid Miner.
Aparna Bhale and Manish Joshi
Springer Singapore
Ravindra Vaidya and Manish Joshi
Springer Singapore
Theyazn H. H. Aldhyani and Manish R. Joshi
Springer Singapore
Swapnaja Gadre-Patwardhan, Vivek Katdare, and Manish Joshi
Springer Singapore
Theyazn H. H. Aldhyani and Manish R. Joshi
IEEE
The network traffic forecasting is of significant interest in many domains such as bandwidth allocation, congestion control and network management. Hence, forecasting of network traffic has received attention from the computer networks field for achieving guaranteed Quality of Service (QoS) in network. In this paper, we propose a forecasting model that combines conventional time series models with clustering approaches. The conventional linear and non linear time series models namely Weighted Exponential Smoothing (WES), Holt-Trend Exponential Smoothing (HTES), AutoRegressive Moving Average (ARMA), Hybrid model (Wavelet with WES) and AutoRegrssive Neural Network (NARNET) models are applied for forecasting network traffic. Our novelty is application of soft clustering for enhancing the existing time series models that are used to forecast network traffic. Clustering can model network traffic data and its characteristics. We derived a methodology to appropriately use cluster centriods to enhance the results obtained by conventional approach. We experimented with different soft clustering techniques such as Fuzzy C-Means (FCM) and Rough K-Means (RKM) clustering to verify the improvement in forecasting. The results of our integrated model are validated using Mean Square Error (MSE), Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE) performance measures. The results show that the integrated model enhances the results obtained using conventional time series forecasting models.
Anurag Bhatt, Sanjay Kumar Dubey, Ashutosh Kumar Bhatt, and Manish Joshi
Springer Singapore
Gajendra Wani and Manish Joshi
IEEE
Constraint-based sequential pattern mining algorithms discover sequential patterns among from sequence data and the resultant sequential patterns satisfy a given constraint. For time stamped sequences duration and/or gap constraints can be applied to obtain corresponding constraint-based sequential patterns. One of the shortcomings of existing algorithms is the requirement to pre-specify a time window to generate time constraint-based sequences. Another limitation is that although these sequences can predict about events that would follow each other, intermediate time interval of these sequences is not available. To overcome these issues, we propose to focus on the estimation of an average intermediate time interval between events that follow each other as a sequence. Instead of sliding a pre-specified time window over sequence data, for a combination of any three events we determine how often these events follow each other as a sequence. A minimum support threshold `min_sup' is used to verify if any given 3-sequence is frequent or not. Earlier, we have proposed an algorithm to enlist inter-transaction associations using 2-sequences [22]. We have extended this work to obtain a list of frequent 3-sequences. We obtained time intervals between successive events of sequences and furthermore we have obtained sequences that satisfy range interval constraints too. The results of our experiments on live retail shop data set are presented.
Manish R. Joshi and Yogita S. Patil
IEEE
Clustering is one of the data mining techniques used in a knowledge discovery process. It is assumed that a good representation of data points may yield good clustering results [6]. This paper discusses the effect of the coordinate system on the clustering. In this paper, we propose a density based clustering approach to group objects represented using Polar coordinate system. The experiment is carried out on different datasets. To evaluate the goodness of cluster result we have used internal and external validity measures. In most of the cases, data points representation using conventional Cartesian coordinate system results in better clustering performance as compared to clustering obtained from same data represented using Polar coordinate system. However, it is observed that for certain Data Sets our proposed density based approach on Polar coordinates clustering results outperform conventional approach. Hence, we can conclude that an appropriate representation of data points may yield more appropriate clustering results.
Theyazn H. H. Aldhyani and Manish R. Joshi
ACM Press
In last decade, highest growth in the numbers of computers connected directly to the Internet has gained momentum. Consequently, prediction of network internet traffic becomes very important and that has received more attention from telecommunication network communities. Further, the quality of service of network like network design, management, planning control, and optimization is enhanced by using prediction network models. It has been found in numerous studies that network traffic is predicted using linear and non-linear time series model. In this paper, we propose an integrated model that combines clustering with Weighted Exponential Smoothing (WES) and AutoRegressive Moving Average (ARMA) models to enhance prediction of packets loading volume in the network traffic. Our experimental results show that the proposed model can be an effective way to improve prediction accuracy achieved with help of k-means clustering. Evaluation and comparison between ARMA, WES models against our proposed model is presented.
Theyazn Hassn Hadi and Manish R. Joshi
IEEE
Intrusion detection system (IDS) is of paramount importance in the present network and system security. Intrusion detection can successfully prevent many attempts to crash network and hamper web services by intruders and hackers. The classification data mining approaches are proposed and used effectively for intrusion detection. However, presences of ambiguous data packets which exhibit traits of two or more classes reduce the overall accuracy of classification. In this paper, we demonstrate the use of supervised partition membership preprocessing method to identify ambiguous packets. We propose an integrated model that results in improved classification accuracy by explicitly clustering ambiguous packets to overcome its misclassification. The novelty of our approach lies in use of non-crisp clustering techniques like fuzzy c-means (FCM) and rough k-means (RKM) that can model ambiguity. Further, we also examined whether FCM clustering and RKM clustering can help to determine class of ambiguous packets exactly or approximately. The support vector machine (SVM) and J48 classifiers results obtained on two standard data sets are presented and compared.
Sangeeta Jadhav, Manish Joshi, and Jyoti Pawar
Informa UK Limited
The ancient Indian classical dance form BharataNatyam (BN) can stay alive and dynamic by allowing innovative, experimental ideas. These comprise of a sequence of possible and legitimate dance steps, and it is estimated that using the main body parts, namely head, neck, hands and legs, more than five lakh dance steps can be generated for a single beat. Thus, dance choreography becomes an intensive, creative, and intuitive process. A choreographer has to finalize appropriate dance steps from among millions of possibilities. Though it is not impossible, the human choreographer cannot explore, analyze and remember all these variations among steps because of the large number of available options. Hence, we propose to develop an autoenumeration followed by autoclassification of significant BN dance steps that can be used in dance performance and choreography. The foremost and most challenging task is to have a computational model that represents different BN dance poses. The second task is to develop a genetic algorithm (GA)-driven automatic system that would provide choreographers a list of unexplored, novel dance steps to fit in a single beat. We designed Art to SMart as a system to model the dance art of BharataNatyam. This system generates dance poses. Furthermore, we have developed a stick figure generation module to help visualize the 30-attribute dance vector generated from the system. The results are evaluated using a mean opinion score measure.
Varsha M. Pathak and Manish R. Joshi
Springer International Publishing
Aparna Bhale, Manish Joshi, and Yogita Patil
Springer International Publishing
Sangeeta Jadhav, Anwaya Aras, Manish Joshi, and Jyoti Pawar
ACM Press
BharataNatyam (BN) is an ancient Indian Classical Dance dating centuries ago. This unique classical dance has been taught by a teacher to a student mostly by rote learning method. A student uses various methods to record the dance choreography taught by a teacher. Currently recording through mobile phones and various other devices of a live dance performance are popular methods but it has its own inherent disadvantages.
Several attempts of automation in choreography are reported and dance visualization is the key factor. Stick Figure representation is a popular method amongst all and is still being used by many of the practitioners.
We have developed a model to represent a BN dance step through a unique thirty attribute dance position vector. Evolutionary approach of Genetic Algorithms generates non-conventional BN dance poses which are approved by renowned dance experts. However, in order to visualize every resulting BN dance, a human model has to pose accordingly. We overcame this hurdle by developing a stick figure generation module. In this paper we present the details of stick figure and in particular how we mold it to suit to a BN dance pose that corresponds to a Dance position vector.
Manish Joshi, Pawan Lingras, Gajendra Wani, and Peng Zhang
IGI Global
This chapter exemplifies how clustering can be a versatile tool in real life applications. Optimal inventory prediction is one of the important issues faced by owners of retail chain stores. Researchers have made several attempts to develop a generic forecasting model for accurate inventory prediction for all products. Regression analysis, neural networks, exponential smoothing, and Autoregressive Integrated Moving Average (ARIMA) are some of the widely used time series prediction techniques in inventory management. However, such generic models have limitations. The authors propose an approach that uses time series clustering and time series prediction techniques to forecast future demand for each product in an inventory management system. A stability and seasonality analysis of the time series is proposed to identify groups of products (local groups) exhibiting similar sales patterns. The details of the experimental techniques and results for obtaining optimal inventory predictions are shared in this chapter.