Distributed Denial of Service (DDoS) attacks are the foremost security concerns on the Internet. DDoS attacks and a similar occurrence called Flash Event (FE) signify anomalies in the normal network... Show More

Distributed Denial of Service (DDoS) attacks are the foremost security concerns on the Internet. DDoS attacks and a similar occurrence called Flash Event (FE) signify anomalies in the normal network traffic, requiring intelligent interventions. This study presents the design and implementation of an intelligent model for the detection of application-layer DDoS attacks and the prevention of service degradations during FE. A Multi-Layer Perceptron (MLP) classifier was used for detecting DDoS attacks on application servers. The FE management system consists of asynchronous processing of requests on a First-In, First-Out (FIFO) basis. A demo application was set up wherein HTTP flood attack was launched and a Flash Event was simulated. The experimental results clearly show that the MLP classifier in comparison with other machine learning classifiers performs best in terms of speed and accuracy. Also, the evaluation of the FE management system shows a great reduction in service degradation. This reflects that the designed model is capable of averting service unavailability on the web.

The processing of big data across different axes is becoming more and more difficult and the introduction of the Hadoop MapReduce framework seems to be a solution to this problem. With this... Show More

The processing of big data across different axes is becoming more and more difficult and the introduction of the Hadoop MapReduce framework seems to be a solution to this problem. With this framework, large amounts of data can be analyzed and processed. It does this by distributing computing tasks between a group of virtual servers operating in the cloud or a large group of devices. The mining process forms an important bridge between data mining and business process analysis. Its techniques make it possible to extract information from event reports. The extraction process generally consists of two phases: identification or discovery and innovation or education. Our first task is to extract small patterns from the log effects. These templates represent the implementation of the tracking from a business process report file. In this step we use the available technologies. Patterns are represented by finite state automation or regular expressions. And the final model is a combination of just two different styles.

The paper presents a complex simulation system for demonstrating the evacuation process in a building, whereby people attempt to escape from a dangerous scenario. It is novel in that it integrates a... Show More

The paper presents a complex simulation system for demonstrating the evacuation process in a building, whereby people attempt to escape from a dangerous scenario. It is novel in that it integrates a range of different models: agent-based model, social force model, and psychological behaviour with emotions and norms. The method uses the communication network based on the message queuing telemetry transport protocol that assists to gather information from the environment. The paths are modified using feelings and rule-based expert system. The authors conduct some simulations and conclude with recommendations for management of safer environments.

Sentiment analysis is a sub-field of natural language processing (NLP). In sentiment analysis the sentiment behind the piece of data is tried to know, this data can be a review of a product by a... Show More

Sentiment analysis is a sub-field of natural language processing (NLP). In sentiment analysis the sentiment behind the piece of data is tried to know, this data can be a review of a product by a customer or a comment on some social media platform. Analysing large amounts of data is still an easy task for small retail websites and business owners. Deep learning (DL) has made a great revolution in the field of speech and image recognition. Mature deep learning neural network i.e. convolution neural network (CNN) has completely changed the field of NLP. This paper proposed a high accuracy, efficient, scalable, reliable and secure solution to cater all the needs of business owners and institutes for sentiment analysis with DL model, a browser based GUI interface for easy accessibility to all the non-technical folks and a dashboard having graphical representations of their results. The proposed sentiment analysis based model has achieved 93.55% accuracy which has outperformed other models.

Named entity recognition is an important factor that has a direct and significant impact on the quality of neural sequence labelling. It entails choosing encoding input data to create grammatical... Show More

Named entity recognition is an important factor that has a direct and significant impact on the quality of neural sequence labelling. It entails choosing encoding input data to create grammatical and semantic representation vectors. The main goal of this research is to provide a hybrid neural network model for a specific sequence labelling task such as named entity recognition. Three subnetworks are used in this hybrid model to ensure that information at the character, capitalization levels, and word-level contextual representation is fully utilized. The authors used different samples for training and development sets on the CoNLL-2003 dataset to show that the model could compare its performance to that of other state-of-the-art models.

A deep neural network is used to develop a covert communication and textual data extraction strategy based on steganography and picture compression in such work. The original input textual image and... Show More

A deep neural network is used to develop a covert communication and textual data extraction strategy based on steganography and picture compression in such work. The original input textual image and cover image are both pre-processed before the covert text-based pictures are separated and implanted into the least significant bit of the cover object picture element using spatial steganography. Following that, stego-images are compressed and transformed(by using Leh Transformation) to provide a higher-quality image while also saving storage space at the sender's end. After then, the stego-image will be transmitted to the receiver over a communication link. At the receiver's end, steganography and compression are then reversed. This work contains a plethora of issues, making it an intriguing subject to pursue. The most crucial component of this task is choosing the right steganography and picture compression technology. The proposed technology, which combines picture steganography with compression and transformation, delivers higher peak signal-to-noise efficiency.

The emerging 5G telecommunication technology uses novel aspects to fulfill the challenges of high data rate, ultra-low latency, broad bandwidth with the best user experience for text detection in... Show More

The emerging 5G telecommunication technology uses novel aspects to fulfill the challenges of high data rate, ultra-low latency, broad bandwidth with the best user experience for text detection in sign board and thereafter transmission of identified information to the vehicles. This is performed on the images which are amorphous in nature or containing scenarios which are random or that cannot be determined. Detecting and transmission of textsover 5G wireless network from the unstructured images aids in many of the additional applications like Optical Character Recognition (OCR) and 5G technolog such as an eMBB, mMTC, and URLLC for quality of service and customer satisfaction.This approach can be used to alert a driver about any road sign even from a captured video by using 5G wireless network irrespective of the weathercondition or any obstacle which may make sign boards difficult to see for drivers. The algorithm uses Maximally Stable Extremal Regions (MSER) feature detector. The algorithm contains several steps which are briefly described in the paper.

One of the major environmental challenges is forest fires, each year millions of hectares of forest are destroyed throughout the world, resulting in economic and ecological damages, as well as the... Show More

One of the major environmental challenges is forest fires, each year millions of hectares of forest are destroyed throughout the world, resulting in economic and ecological damages, as well as the loss of human life. Therefore, predicting forest fires is of great importance for governments; However, there is still limited study on this topic in Algeria. In this paper, we present an application of artificial neural networks to predict forest fires in embedded devices. We used meteorological data obtained from wireless sensor networks. In the experimentation, nine machine learning model are compared. The findings from this study make several contributions to the current literature. First, our model is suitable for embedded and real-time training and prediction. Moreover, it should provide better performances and accurate predictions against other models.

India as a country has 17.7% of the world’s population with the limited availability of land resource which is about only 2.4% of the world’s land. Being a developing nation and such huge population... Show More

India as a country has 17.7% of the world’s population with the limited availability of land resource which is about only 2.4% of the world’s land. Being a developing nation and such huge population to accommodate, a number of problems can be seen on a daily basis such as high traffic congestion and unmanaged traffic on the roads. Irritating rush, wastage of time and fuel, are being severe hindrance to make the transportation comfortable. As a country, due to availability of limited lands, the only option is to manage the traffic smartly. Hitherto, a number of attempts have been made in this regard, still the statically managed traffic lights can be seen at the junction of roads. So in this work, it was tried to give an easy, but implementable method to manage traffic lights effectively. A hybrid approach based enhanced Convolution Neural Network model was used for the classification and have given the comparison with other model based technique i.e. Support Vector Machine. Our proposed enhanced model produced 91.01% accuracy and it is able to outperform the existing model.

Complex computational problems are occurrences in our daily lives that needs to be analysed effectively in order to make meaningful and informed decision. This study performs empirical analysis... Show More

Complex computational problems are occurrences in our daily lives that needs to be analysed effectively in order to make meaningful and informed decision. This study performs empirical analysis into the performance of six optimisation algorithms based on swarm intelligence on nine well known stochastic and global optimisation problems, with the aim of identifying a technique that returns an optimum output on some selected benchmark techniques. Extensive experiments show that, Multi-Swarm and Pigeon inspired optimisation algorithm outperformed Particle Swarm, Firefly and Evolutionary optimizations in both convergence speed and global solution. The algorithms adopted in this paper gives an indication of which algorithmic solution presents optimal results for a problem in terms of quality of performance, precision and efficiency.

In our daily life, it is difficult to meet financial demand while in crisis. This financial crisis may be solved with financial assistance from the banks. The financial assistance is nothing but... Show More

In our daily life, it is difficult to meet financial demand while in crisis. This financial crisis may be solved with financial assistance from the banks. The financial assistance is nothing but availing loan from the bank with proper agreement to repay the amount including calculated interest within the loan approved tenure. The customer can only avail loans against the submission of some valid and important supportive documents. However, although the customer is aware of the whole process of repayment and installment along with loan approval tenure, most of the time it is hard to get the approved loan within a shorter period. Therefore, the objective of this paper is to automate this manual and long process by predicting the chance of approval of the loan. The novelty of this research article is to apply machine learning techniques and classification algorithms to predict loan eligibility through an automatic online loan application process

This paper examines the new viewpoints that have evolved as a result of the advent of internet of things (IoT) in the kitchen. Companies are now exploring how internal knowledge and skillsets relate... Show More

This paper examines the new viewpoints that have evolved as a result of the advent of internet of things (IoT) in the kitchen. Companies are now exploring how internal knowledge and skillsets relate to the new technical needs that evolving digital environment entails; and they are learning more about IoT and connected products going through internal research. Accordingly, they hope to rely on the internet of things to keep kitchens safe. Cooking leads to cause of house fires and fire injuries. The bulk of the fires in the building started in the kitchens of the units. Three elements are required to start and maintain a fire. The human body is made up of these three elements: fuel, heat, and oxygen. Fire safety measures includes protect building from damage and death. An IoT-based system detects CO2 and Methane (CH4) levels in the environment and kitchen, as well as temperature. It has the ability to prevent accidents and save lives and property. When sensor data is synced, an IoT-based controlling device sends notifications to the mobile phones of the chosen number set in the alert section.

Detection of humans in flames is a challenging task. The task in this work is classified into two stages. The first is detection of fire, and the second is detection of human. The proposed method... Show More

Detection of humans in flames is a challenging task. The task in this work is classified into two stages. The first is detection of fire, and the second is detection of human. The proposed method involves fire detection based on colour format YCbCr for image preprocessing. It further uses a histogram of oriented gradient (HOG) and support vector machine (SVM) to detect a human in the fire. It evaluates several motion-based feature sets for human detection in the form of videos. In this work, both modules were integrated to make them work together. For the detection of fire, four different rules involving colour thresholding were used and background differencing was used for moving object detection. The main objective of this work is to spot the humans in the flames who are trapped in it so they can be rescued quickly. This can help the firefighters in rapid planning and serious zone detection. The proposed model has 81% efficiency, which has outperformed the existing models for detection of humans in flames.

Object detection is used in almost every real-world application such as autonomous traversal, visual system, face detection, and even more. This paper aims at applying object detection technique to... Show More

Object detection is used in almost every real-world application such as autonomous traversal, visual system, face detection, and even more. This paper aims at applying object detection technique to assist visually impaired people. It helps visually impaired people to know about the objects around them to enable them to walk free. A prototype has been implemented on a Raspberry PI3 using OpenCV libraries, and satisfactory performance is achieved. In this paper, a detailed review has been carried out on object detection using region-conventional neural network (RCNN)-based learning systems for a real-world application. This paper explores the various process of detecting objects using various object detections methods and walks through detection including a deep neural network for SSD implemented using Caffee model.