Abstract Today is the Golden Age of Hacking. Any person with malicious intents can acquire tools ... more Abstract Today is the Golden Age of Hacking. Any person with malicious intents can acquire tools and techniques via numerous freely hosted sites to launch attacks on Networks. Identifying and eliminating security threats has become an arduous task for the ...
A massive amount of textual data now exists in digital repositories in the form of research artic... more A massive amount of textual data now exists in digital repositories in the form of research articles, news articles, reviews, Wikipedia articles, and books, etc. Text clustering is a fundamental data mining technique to perform categorization, topic extraction, and information retrieval. Textual datasets, especially which contain a large number of documents are sparse and have high dimensionality. Hence, traditional clustering techniques such as K-means, Agglomerative clustering, and DBSCAN cannot perform well. In this paper, a clustering technique especially suitable to large text datasets is proposed that overcome these limitations. The proposed technique is based on word embeddings derived from a recent deep learning model named “Bidirectional Encoders Representations using Transformers”. The proposed technique is named as WEClustering. The proposed technique deals with the problem of high dimensionality in an effective manner, hence, more accurate clusters are formed. The techni...
Journal of Ambient Intelligence and Humanized Computing
Transforming text from one language to another by using computer systems automatically or with li... more Transforming text from one language to another by using computer systems automatically or with little human interventions is known as Machine Translation System (MTS). Divergence among natural languages in a multilingual environment makes Machine Translation (MT) a difficult and challenging task. The purpose of this paper is to present a comprehensive survey of MTS in general and for English, Hindi and Sanskrit languages in particular. The state-of-the-art MT approach is Neural Machine Translation (NMT) which has been used by Google, Amazon, Facebook and Microsoft but it requires large corpus as well as high computing systems. The availability of MT language modeling tools, parsers data repositories and evaluation metrics has been tabulated in this article. The classification of MTS, evaluation methods and platforms has been done based on a well-defined set of criteria. The new research avenues have been explored in this survey article which will help in developing good quality MTS. Although several surveys have been done on MTS but none of them have followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) approach including tools and evaluation methods as done in this survey specifically for English, Hindi and Sanskrit languages.
Abstract Now-a-days, with the ever increasing number of vehicles, getting parking space at right ... more Abstract Now-a-days, with the ever increasing number of vehicles, getting parking space at right place and on time has become an inevitable necessity for all across the globe. In this context, finding an unoccupied parking slot by the interested vehicle owners with least overhead becomes an NP-Hard problem bounded by various constraints. In-advance availability of information regarding parking occupancy plays a major role in hassle free trip optimization for motorists. It also facilitates services-cum-profit management for the parking owners. It further helps in curbing congestion by reducing cruising time and hence, helps in controlling pollution of the smart cities. Thus, accurate and timely information regarding parking occupancy and availability has become the basic need in the evolution of the smart cities. Motivated by these facts, an occupancy-driven machine learning based on-street parking pricing scheme is proposed in this paper. The proposed scheme uses machine learning based approaches to predict occupancy of parking lots, which in turn is used to deduce occupancy driven prices for arriving vehicles. In order to train, test, and compare different machine learning models, on-street parking data of Seattle city has been used. To the best of our knowledge, this is the first time that parking occupancy prediction system is used to generate occupancy based parking prices for on-street parking system of the Seattle city. Results obtained using the proposed occupancy driven machine learning based on-street parking pricing scheme demonstrate its effectiveness over other existing state-of-the-art schemes.
ABSTRACT This paper is a case study of utilizing machine learning for developing a decision-makin... more ABSTRACT This paper is a case study of utilizing machine learning for developing a decision-making system for auditors before initializing the audit fieldwork of public firms. Annual data of 777 firms from 14 different sectors are collected and a MCTOPE (Multi criteria ToPsis based Ensemble) framework is implemented to build an ensemble classifier. MCTOPE framework optimizes the performance of classification during ensemble building using the TOPSIS multi-criteria decision-making algorithm. Ensemble machine learning is used for optimizing the prediction performance of suspicious firm predictor in the previous work available at https://www.tandfonline.com/doi/full/10.1080/08839514.2018.1451032. After achieving an accuracy of 94.6% and AUC (area under the curve) value of 0.98, this ensemble classifier is employed in a web application developed for auditors using Python and R script for the prediction of suspicious firm before planning an external audit. The performance of an ensemble classifier is validated using K-fold cross validation technique and is found to be better than the state-of-the-art classifiers.
Vehicular ad hoc networks (VANETs) have emerged as fast growing networks with aims to provide saf... more Vehicular ad hoc networks (VANETs) have emerged as fast growing networks with aims to provide safety, and comfort to the onboard passengers. But, in this environment reliable data dissemination to the destination nodes is one of the biggest challenges as there may be a congestion in the network due to blind flooding of messages to their final destination, called as broadcast storm which may lead to the performance degradation with respect to the metric such as-message delivery, reliability and response time. To address this issue, in this paper, we propose a reliability-aware intelligent data dissemination protocol for broadcast storm problem in VANETs. We have solved the above specified problem using game theory concepts in which players, strategy space and decisions with respect to the current state of the system. To reduce the message overhead with respect to the communication cost among the players of the game, in the proposed scheme, messages are routed to the next destination by selecting the most reliable path in an intelligent manner. A coalition game is formulated among the vehicles by considering vehicles as the players in the game. Each player in the game has an initial payoff value based upon the parameters such as-communication range, storage requirements, and computation power. Based upon the payoff value of the players in the game, different coalitions/clusters are formulated among the players in the proposed scheme. Then, the players send the messages to the other vehicles/players within the same coalition which increases the reliability of transmission. In the proposed scheme, messages are unicasted to their final destination in an intelligent manner. Different algorithms are designed for coalition formation, maintenance, and reliability-aware data dissemination. The performance of the proposed scheme is evaluated using various evaluation metrics such as-service time, packet delivery ration, and throughput. The results obtained prove the effectiveness of the proposed scheme as compared to the other state-of-the-art existing schemes.
International Journal of Ad Hoc and Ubiquitous Computing
In universities, IT infrastructure is usually non-centralised. Setting up a private cloud setup i... more In universities, IT infrastructure is usually non-centralised. Setting up a private cloud setup in academic institutes has an edge over the traditional approach of IT infrastructure distribution to its various departments, schools and centres etc. This work proposes a energy efficient resource management solution specially designed for small and medium sized Infrastructure as a Service IaaS cloud in Universities. The proposed algorithm optimises the energy consumption through efficient resource scheduling, virtual machine VM consolidation and VM admission control policies. Proposed algorithm has been evaluated by keeping in consideration the number of accepted incoming VM requests and energy consumption of cloud infrastructure. Experimental results reveal that proposed algorithm enhances the acceptance rate of VM requests with the reduction of overall energy consumption of cloud infrastructure.
Various indexing methods of spatial data have come out after rigorous e↵orts put by many research... more Various indexing methods of spatial data have come out after rigorous e↵orts put by many researchers for fast processing of spatial queries. Parallelizing spatial index building and query processing have become very popular for improving eciency. The MapReduce framework provides a modern way of parallel processing. A MapReduce-based works for spatial queries consider the existing traditional spatial indexing for building spatial indexes in parallel. The majority of the spatial indexes implemented in MapReduce use R-Tree and its variants. Therefore, R-Tree and its variant-based traditional spatial indexes are thoroughly surveyed in the paper. The objective is to search for still less explored spatial indexing approaches, having the potential for par-allelism in MapReduce. The review work also provides a detailed survey of MapReduce-based spatial query processing approaches-hierarchical indexed and packed key-value storage based spatial dataset. Both approaches use di↵er-ent data part...
2014 Ieee International Conference on Cloud Computing in Emerging Markets, Oct 1, 2014
Nowadays, with the increased deployment of servers to facilitate high performance computing (HPC)... more Nowadays, with the increased deployment of servers to facilitate high performance computing (HPC) for scientific and engineering applications lead to large consumption of energy. Cloud computing is a cost-effective solution, as it allows to host storage, computational and supported network services on a shared infrastructure of physical servers. However, the growing demand of cloud infrastructure among the IT companies is drastically increasing, by which data centers are drawing more energy. Energy efficient scheduling is one effective solution to streamline the resource usage as well as reduce the energy consumption. The proposed work in this paper demonstrates the resource allocation and makes an energy consumption analysis of Greedy, Round Robin and Power Aware Best Fit Decreasing scheduling algorithms on a private academic cloud. This paper provides an insight into the working of different scheduling scenarios for cloud computing and demonstrates the potential for the improvement of energy efficiency of PABFD algorithm under academic workload.
Abstract Today is the Golden Age of Hacking. Any person with malicious intents can acquire tools ... more Abstract Today is the Golden Age of Hacking. Any person with malicious intents can acquire tools and techniques via numerous freely hosted sites to launch attacks on Networks. Identifying and eliminating security threats has become an arduous task for the ...
A massive amount of textual data now exists in digital repositories in the form of research artic... more A massive amount of textual data now exists in digital repositories in the form of research articles, news articles, reviews, Wikipedia articles, and books, etc. Text clustering is a fundamental data mining technique to perform categorization, topic extraction, and information retrieval. Textual datasets, especially which contain a large number of documents are sparse and have high dimensionality. Hence, traditional clustering techniques such as K-means, Agglomerative clustering, and DBSCAN cannot perform well. In this paper, a clustering technique especially suitable to large text datasets is proposed that overcome these limitations. The proposed technique is based on word embeddings derived from a recent deep learning model named “Bidirectional Encoders Representations using Transformers”. The proposed technique is named as WEClustering. The proposed technique deals with the problem of high dimensionality in an effective manner, hence, more accurate clusters are formed. The techni...
Journal of Ambient Intelligence and Humanized Computing
Transforming text from one language to another by using computer systems automatically or with li... more Transforming text from one language to another by using computer systems automatically or with little human interventions is known as Machine Translation System (MTS). Divergence among natural languages in a multilingual environment makes Machine Translation (MT) a difficult and challenging task. The purpose of this paper is to present a comprehensive survey of MTS in general and for English, Hindi and Sanskrit languages in particular. The state-of-the-art MT approach is Neural Machine Translation (NMT) which has been used by Google, Amazon, Facebook and Microsoft but it requires large corpus as well as high computing systems. The availability of MT language modeling tools, parsers data repositories and evaluation metrics has been tabulated in this article. The classification of MTS, evaluation methods and platforms has been done based on a well-defined set of criteria. The new research avenues have been explored in this survey article which will help in developing good quality MTS. Although several surveys have been done on MTS but none of them have followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) approach including tools and evaluation methods as done in this survey specifically for English, Hindi and Sanskrit languages.
Abstract Now-a-days, with the ever increasing number of vehicles, getting parking space at right ... more Abstract Now-a-days, with the ever increasing number of vehicles, getting parking space at right place and on time has become an inevitable necessity for all across the globe. In this context, finding an unoccupied parking slot by the interested vehicle owners with least overhead becomes an NP-Hard problem bounded by various constraints. In-advance availability of information regarding parking occupancy plays a major role in hassle free trip optimization for motorists. It also facilitates services-cum-profit management for the parking owners. It further helps in curbing congestion by reducing cruising time and hence, helps in controlling pollution of the smart cities. Thus, accurate and timely information regarding parking occupancy and availability has become the basic need in the evolution of the smart cities. Motivated by these facts, an occupancy-driven machine learning based on-street parking pricing scheme is proposed in this paper. The proposed scheme uses machine learning based approaches to predict occupancy of parking lots, which in turn is used to deduce occupancy driven prices for arriving vehicles. In order to train, test, and compare different machine learning models, on-street parking data of Seattle city has been used. To the best of our knowledge, this is the first time that parking occupancy prediction system is used to generate occupancy based parking prices for on-street parking system of the Seattle city. Results obtained using the proposed occupancy driven machine learning based on-street parking pricing scheme demonstrate its effectiveness over other existing state-of-the-art schemes.
ABSTRACT This paper is a case study of utilizing machine learning for developing a decision-makin... more ABSTRACT This paper is a case study of utilizing machine learning for developing a decision-making system for auditors before initializing the audit fieldwork of public firms. Annual data of 777 firms from 14 different sectors are collected and a MCTOPE (Multi criteria ToPsis based Ensemble) framework is implemented to build an ensemble classifier. MCTOPE framework optimizes the performance of classification during ensemble building using the TOPSIS multi-criteria decision-making algorithm. Ensemble machine learning is used for optimizing the prediction performance of suspicious firm predictor in the previous work available at https://www.tandfonline.com/doi/full/10.1080/08839514.2018.1451032. After achieving an accuracy of 94.6% and AUC (area under the curve) value of 0.98, this ensemble classifier is employed in a web application developed for auditors using Python and R script for the prediction of suspicious firm before planning an external audit. The performance of an ensemble classifier is validated using K-fold cross validation technique and is found to be better than the state-of-the-art classifiers.
Vehicular ad hoc networks (VANETs) have emerged as fast growing networks with aims to provide saf... more Vehicular ad hoc networks (VANETs) have emerged as fast growing networks with aims to provide safety, and comfort to the onboard passengers. But, in this environment reliable data dissemination to the destination nodes is one of the biggest challenges as there may be a congestion in the network due to blind flooding of messages to their final destination, called as broadcast storm which may lead to the performance degradation with respect to the metric such as-message delivery, reliability and response time. To address this issue, in this paper, we propose a reliability-aware intelligent data dissemination protocol for broadcast storm problem in VANETs. We have solved the above specified problem using game theory concepts in which players, strategy space and decisions with respect to the current state of the system. To reduce the message overhead with respect to the communication cost among the players of the game, in the proposed scheme, messages are routed to the next destination by selecting the most reliable path in an intelligent manner. A coalition game is formulated among the vehicles by considering vehicles as the players in the game. Each player in the game has an initial payoff value based upon the parameters such as-communication range, storage requirements, and computation power. Based upon the payoff value of the players in the game, different coalitions/clusters are formulated among the players in the proposed scheme. Then, the players send the messages to the other vehicles/players within the same coalition which increases the reliability of transmission. In the proposed scheme, messages are unicasted to their final destination in an intelligent manner. Different algorithms are designed for coalition formation, maintenance, and reliability-aware data dissemination. The performance of the proposed scheme is evaluated using various evaluation metrics such as-service time, packet delivery ration, and throughput. The results obtained prove the effectiveness of the proposed scheme as compared to the other state-of-the-art existing schemes.
International Journal of Ad Hoc and Ubiquitous Computing
In universities, IT infrastructure is usually non-centralised. Setting up a private cloud setup i... more In universities, IT infrastructure is usually non-centralised. Setting up a private cloud setup in academic institutes has an edge over the traditional approach of IT infrastructure distribution to its various departments, schools and centres etc. This work proposes a energy efficient resource management solution specially designed for small and medium sized Infrastructure as a Service IaaS cloud in Universities. The proposed algorithm optimises the energy consumption through efficient resource scheduling, virtual machine VM consolidation and VM admission control policies. Proposed algorithm has been evaluated by keeping in consideration the number of accepted incoming VM requests and energy consumption of cloud infrastructure. Experimental results reveal that proposed algorithm enhances the acceptance rate of VM requests with the reduction of overall energy consumption of cloud infrastructure.
Various indexing methods of spatial data have come out after rigorous e↵orts put by many research... more Various indexing methods of spatial data have come out after rigorous e↵orts put by many researchers for fast processing of spatial queries. Parallelizing spatial index building and query processing have become very popular for improving eciency. The MapReduce framework provides a modern way of parallel processing. A MapReduce-based works for spatial queries consider the existing traditional spatial indexing for building spatial indexes in parallel. The majority of the spatial indexes implemented in MapReduce use R-Tree and its variants. Therefore, R-Tree and its variant-based traditional spatial indexes are thoroughly surveyed in the paper. The objective is to search for still less explored spatial indexing approaches, having the potential for par-allelism in MapReduce. The review work also provides a detailed survey of MapReduce-based spatial query processing approaches-hierarchical indexed and packed key-value storage based spatial dataset. Both approaches use di↵er-ent data part...
2014 Ieee International Conference on Cloud Computing in Emerging Markets, Oct 1, 2014
Nowadays, with the increased deployment of servers to facilitate high performance computing (HPC)... more Nowadays, with the increased deployment of servers to facilitate high performance computing (HPC) for scientific and engineering applications lead to large consumption of energy. Cloud computing is a cost-effective solution, as it allows to host storage, computational and supported network services on a shared infrastructure of physical servers. However, the growing demand of cloud infrastructure among the IT companies is drastically increasing, by which data centers are drawing more energy. Energy efficient scheduling is one effective solution to streamline the resource usage as well as reduce the energy consumption. The proposed work in this paper demonstrates the resource allocation and makes an energy consumption analysis of Greedy, Round Robin and Power Aware Best Fit Decreasing scheduling algorithms on a private academic cloud. This paper provides an insight into the working of different scheduling scenarios for cloud computing and demonstrates the potential for the improvement of energy efficiency of PABFD algorithm under academic workload.
A WSN may be considered as a distributed database because of the presence of data in physically d... more A WSN may be considered as a distributed database because of the presence of data in physically dispersed nodes constituting it. Data is extracted by various applications to serve the information needs. A number of techniques are being used or being introduced to extract data. Data extraction through queries is the most popular approach due to its ease of use. To combat the various limitations of limited power, bandwidth, node failure rate researchers are devising variants of querying interfaces. Main thrust is to achieve energy efficiency.
Uploads