Browse Items (1422 total)
Sort by:
-
Therapy recommendation based on level of depression using social media data
Social media is a massive platform with currently over 100 million registered users. It is a platform where individuals express themselves along with their interests. These expressions of individual can be used to identify their mental status. That being said, depression and anxiety are the dominant cause for illness and ill-health across the world. Studies show that users mental health can be predicted by their everyday use of language. This paper examines the tweets for analyzing the linguistic and behavioral features for classifying the levels of depression among the users. In order to classify the levels of depression, a knowledge base of the words that are associated with depression/anxiety has been created. The model evaluated this using simple text mining techniques to measure the mental health status of the users and provide appropriate recommendations. The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021. -
Wearable Sensors for Pervasive and Personalized Health Care
Healthcare systems are designed to provide commendable services to cater health needs of individuals with minimum expenditure and limited use of human resources. Pervasive health care can be considered as a major development in the healthcare system which aims to treat patients with minimal human resources. This provides a solution to several existing healthcare problems which might change the future of the healthcare systems in a positive way. Pervasive health care is defined as a system which is available to anyone at any point of time and at any place without any location constraints. At a broader definition, it helps in monitoring the health-related issues at a home-based environment by medical stakeholders which is very beneficial in case of emergency situations. This chapter elaborates architecture of IoT, how wearable sensors can be used to help people to get personalized and pervasive healthcare systems, and it also gives a detailed working of different types of IoT-enabled wearable devices for pervasive health care. 2021, The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
On some classes of equitable irregular graphs
Graph labeling techniques are used by data scientists to represent data points and their relationships with each other. The segregation/sorting of similar datasets/points are easily done using labeling of vertices or edges in a graph. An equitable irregular edge labeling is a function $$f: E(G) \rightarrow N$$ (not necessarily be injective) such that the vertex sums of any two adjacent vertices of $$G$$ differ by at most one, where vertex sum of a vertex is the sum of the labels under $$f$$ of the edges incident with that vertex. A graph admitting an equitable irregular edge labeling is called an equitable irregular graph (EIG). In this paper, more classes of equitable irregular graphs are presented. We further generalize the concept of equitable irregular edge labeling to $$k$$-equitable irregular edge labeling by demanding the difference of the vertex sum of adjacent vertices to be $$k \ge 1$$. The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021. -
Adaptive artificial bee colony (aabc)-based malignancy pre-diagnosis
Lung cancer is one of the leading causes of death. The survival rate of the patients diagnosed with lung cancer depends on the stage of the detection and the timely prognosis. Hence, early detection of anomalous malignant cells is needed for pre-diagnosis of lung cancer as it plays a major role in the prognosis and treatment. In this work, an innovative pre-diagnosis approach is suggested, wherein the size of the dataset comprising risk factors and symptoms is considerably decreased and optimized by means of an Adaptive Artificial Bee Colony (AABC) algorithm. Subsequently, the optimized dataset is fed to the Feed-Forward Back-Propagation Neural Network (FFBNN) to perform the training task. For the testing, supplementary data is furnished to well-guided FFBNN-AABC to authenticate whether the supplied investigational data is competent to effectively forecast the lung disorder or not. The results obtained show a considerable improvement in the classification performance compared to other approaches like Genetic Algorithm (GA) and Particle Swarm Optimization (PSO). The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021. -
A comparative study of text mining algorithms for anomaly detection in online social networks
Text mining is a process by which information and patterns are extracted from textual data. Online Social Networks, which have attracted immense attention in recent years, produces enormous text data related to the human behaviours based on their interactions with each other. This data is intrinsically unstructured and ambiguous in nature. The data involves incorrect spellings and inaccurate grammars leading to lexical, syntactic and semantic ambiguities. This causes wrong analysis and inappropriate pattern identification. Various Text Mining approaches are being used by researchers which can help in Anomaly Detection through Topic Modeling, identification of Trending Topics, Hate Speeches and evolution of the communities in Online Social Networks. In this paper, a comparative analysis of the performance of four classification algorithms, Support Vector Machine (SVM), Rocchio, Decision Trees and K-Nearest Neighbour (KNN) for a Twitter data set is presented. The experimental study revealed that SVM outperforms better than other classifiers, and also classifies the dataset into anomalous and non-anomalous users opinions. The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021. -
A deep learning approach in early prediction of lungs cancer from the 2d image scan with gini index
Digital Imaging and Communication in Medicine (DiCoM) is one of the key protocols for medical imaging and related data. It is implemented in various healthcare facilities. Lung cancer is one of the leading causes of death because of air pollution. Early detection of lung cancer can save many lives. In the last 5years, the overall survival rate of lung cancer patients has increased, due to early detection. In this paper, we have proposed Zero-phase Component Analysis (ZCA) whitening and Local Binary Pattern (LBP) to enhance the quality of lung images which will be easy to detect cancer cells. Local Energy based Shape Histogram (LESH) technique is used to detect lung cancer. LESH feature extracts a suitable diagnosis of cancer from the CT scans. The Gini coefficient is used for characterizing lung nodules which will be helpful in Computed Tomography (CT) scan. We propose a Convolutional Neural Network (CNN) algorithm to integrate multilayer perceptron for image segmentation. In this process, we combined both traditional feature extraction and high-level feature extraction to classify lung images. The convolutional neural network for feature extraction will identify lung cancer cells with traditional feature extraction and high-level feature extraction to classify lung images. The experiment showed a final accuracy of about 93.27%. The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021. -
Treexpan instantiation of xpattern framework
Most of the data generated from social media, Internet of Things, etc. are semi-structured or unstructured. XML is a leading semi-structured data commonly used over cross-platforms. XML clustering is an active research area. Because of the complexity of XML clustering, it remains a challenging area in data analytics, especially when Big Data is considered. In this paper, we focus on clustering of XML based on structure. A novel method for representing XML documents, Compressed Representation of XML Tree, is proposed following the concept of frequent pattern tree structure. From the proposed structure, clustering is carried out with a new algorithm, TreeXP, which follows the XPattern framework. The performances of the proposed representation and clustering algorithm are compared with a well-established PathXP algorithm and found to give the same performance, but require very less time. The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021. -
Implementation of integer factorization algorithm with pisano period
The problem of factorization of large integers into the prime factors has always been of mathematical interest for centuries. In this paper, starting with a historical overview of integer factorization algorithms, the study is extended to some recent developments in the prime factorization with Pisano period. To reduce the computational complexity of Fibonacci number modulo operation, the fast Fibonacci modulo algorithm has been used. To find the Pisano periods of large integers, a stochastic algorithm is adopted. The Pisano period factorization method has been proved slightly better than the recently developed algorithms such as quadratic sieve method and the elliptic curve method. This paper ideates new insights in the area of integer factorization problems. The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021. -
An advanced machine learning framework for cybersecurity
The world is turning out to be progressively digitalized raising security concerns and the urgent requirement for strong and propelled security innovations and procedures to battle the expanding complex nature of digital assaults. This paper examines how AI is being utilized in digital security in both resistance and offense exercises, remembering exchanges for digital attacks focused on AI models. Digital security is the assortment of approaches, systems, advancements, and procedures that work together to ensure the confidentiality, trustworthiness, and accessibility of processing assets, systems, programming projects, and information from attacks. Machine learning-based examination for cybersecurity is the following rising pattern in digital security, planned for mining security information to reveal progressed focused on digital threats and limiting the operational overheads of keeping up static relationship rules. In this paper, we are mainly focusing on the detection and diagnosis of various cyber threats based on machine learning. The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021. -
Classification of financial news articles using machine learning algorithms
The opinion helps in determining the direction of the stock market. Information hidden in news articles is an information treasure which needs to be extracted. The present study is conducted to explore the application of text mining in binning the financial articles according to the opinion expressed inside them. It is discovered that using the tri-n-gram feature extraction process in conjugation with Support Vector machines increases the reliability and precision of the binning process. The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021. -
Efficient handwritten character recognition of modi script using wavelet transform and svd
MODI script has historical importance as it was used for writing the Marathi language, until 1950. Due to the complex nature of the script, the character recognition of MODI script is still in infancy. The implementation of more efficient methods at the various stages of the character recognition process will increase the accuracy of the process. In this paper, we present a hybrid method called WT-SVD (Wavelet Transform-Singular Value Decomposition), for the character recognition of MODI script. The WT-SVD method is a combination of singular value decomposition and wavelet transform, which is used for the feature extraction. Euclidean distance method is used for the classification. The experiment is conducted using Symlets and Biorthogonal wavelets, and the results are compared. The method using Biorthogonal wavelet feature extraction achieved the highest accuracy The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021. -
A predictive model on post-earthquake infrastructure damage
Disaster management initiatives are employed to mitigate the effects of catastrophic events such as earthquakes. However, post-disaster expenses raise concern for both the government and the insurance companies. The paper provides insights about the key factors that add to the building damage such as the structural and building usage properties. It also sheds light on the best model that can be adopted in terms of both accuracy and ethical principles such as transparency and accountability. From the performance perspective, random forest model has been suggested. From the perspective of models with ethical principles, the decision tree model has been highlighted. Thus, the paper fulfills to propose the best predictive model to accurately predict the building damage caused by earthquake for incorporation by the insurance companies or government agency to minimize the post-disaster expenses involved in such catastrophic event. The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021. -
Body mass index implications using data analysis in the soccer sports
Soccer is considered among the most popular sports in the world among the last few years. At the same time, it has become a prime target in developing countries like India and other Asian countries. As science and technology grow, we can see that sports also grow with science, and hence technology being used to determine the results sometime or sometimes it is used to grow the overall effect. This paper presents the attributes and the qualities which are necessary to develop in a player in order to play for the big-time leagues called Premier League, La Liga, Serie A, German Leagues and so on. Simple correlation and dependence techniques have been used in this paper in order to get proper relationship among the attributes. This paper also examines how the body mass index plays an effect on the presentation of soccer players with respect to their speed, increasing speed, work rate, aptitude moves and stamina. The point is likewise to discover the connection of the above credits concerning body mass index. As in universal exchange, football clubs can profit more in the event that they have practical experience in what they have or can make a similar bit of room to maneuver. In a universe of rare assets, clubs need to recognize what makes them effective and contribute in like manner. Springer Nature Singapore Pte Ltd 2021. -
Analysis of attention deficit hyperactivity disorder using various classifiers
Attention Deficit Hyperactivity Disorder (ADHD) is a neurobehavioral childhood impairment that wipes away the beauty of the individual from a very young age. Data mining classification techniques which are becoming a very important field in every sector play a vital role in the analysis and identification of these disorders. The objective of this paper is to analyze and evaluate ADHD by applying different classifiers like Nae Bayes, Bayes Net, Sequential Minimal Optimization, J48 decision tree, Random Forest, and Logistic Model Tree. The dataset employed in this paper is the first publicly obtainable dataset ADHD-200 and the instances of the dataset are classified into low, moderate, and high ADHD. The analysis of the performance metrics and therefore the results show that the Random Forest classifier offers the highest accuracy on ADHD dataset compared to alternative classifiers. With the current need to provide proper evaluation and management of this hyperactive disorder, this research would create awareness about the influence of ADHD and can help ensure the proper and timely treatment of the affected ones. Springer Nature Singapore Pte Ltd 2021. -
Ocr system framework for modi scripts using data augmentation and convolutional neural network
Character recognition is one of the most active research areas in the field of pattern recognition and machine intelligence. It is a technique of recognizing either printed or handwritten text from document images and converting it to a machine-readable form. Even though there is much advancement in the field of character recognition using machine learning techniques, recognition of handwritten MODI script, which is an ancient Indian script, is still in its infancy. It is due to the complex nature of the script that includes similar shapes of character and the absence of demarcation between words. MODI was an official language used to write Marathi. Deep learning-based models are very efficient in character recognition tasks and in this work an ACNN model is proposed using the on-the-fly data augmentation method and convolution neural network. The augmentation of the data will add variability and generalization to the data set. CNN has special convolution and pooling layers which have helped in better feature extraction of the characters. The performance of the proposed method is compared with the most accurate MODI character recognition method reported so far and it is found that the proposed method outperforms the other method. The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021. -
Tag indicator: a new predictive tool for stock trading
In this paper, TAGan indicator for stock market prediction in which volume-based means for measuring potential trading and investing decision-making is introduced. This task has been in correlation of the changes in the volume with the changes in the actual trade volume. Using this, a concise trading strategy is formulated. Hoping to outperform the market and analyze the results by back testing across intraday, price data for the last 1 year, 2019, is performed. It was discovered that about 48.9% of the time, the volume-based trading strategy outperformed and the returns from market are also healthy enough to support the claim. Statistical methods like linear regression, mean square error in prediction and stochastic gradient descent are applied. Furthermore, while the scope of the study was limited to a few stocks in Nifty in order to mitigate selection bias, nonetheless, we hypothesize that numerous other assets that similarly possess a predictable correlation to volumes based on daily high and low are likely to exist. The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021. -
Diabetic retinopathy detection using convolutional neural networka study
Detection and classification of Diabetic Retinopathy (DR) is a challenging task. Automation of the detection is an active research area in image processing and machine learning. Conventional preprocessing and feature extraction methods followed by classification of a suitable classifier algorithm are the common approaches followed by DR detection. With the advancement in deep learning and the evolution of Convolutional Neural Network (CNN), conventional preprocessing and feature extraction steps are rapidly being replaced by CNN. This paper reviews some of the recent contributions in diabetic retinopathy detection using deep architectures. Further, two architectures are implemented with minor modifications. Experiments are carried out with different sample sizes, and the detection accuracies of the two architectures are compared. The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021. -
Toxic text classification
The users of the Internet increase every moment with increasing population and accessibility of the Internet. With the increase in the number of users of the Internet, the number of controversies, arguments and abuses of all kinds increases. It becomes necessary for social media and other sites to identify toxic content amongst a large number of content being posted by the users of the sites every second. The traditional algorithms that depend on users reporting toxic content for it to be deleted and necessary actions to be taken against the users posting the content would take a long time, within which it would have gained media attention and would have lead to huge fights over the content. Thus, it becomes important for the content to be evaluated for toxicity at the time it is posted in order to stop it from being posted. Therefore, we have designed and trained a deep learning model that can be read through the textual content given through it and determine if it is toxic or not. The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021. -
Algorithms for the metric dimension of a simple graph
Let G = (V, E) be a connected, simple graph with n vertices and m edges. Let v1, v2 $$\in$$ V, d(v1, v2) is the number of edges in the shortest path from v1 to v2. A vertex v is said to distinguish two vertices x and y if d(v, x) and d(v, y) are different. D(v) as the set of all vertex pairs which are distinguished by v. A subset of V, S is a metric generator of the graph G if every pair of vertices from V is distinguished by some element of S. Trivially, the whole vertex set V is a metric generator of G. A metric generator with minimum cardinality is called a metric basis of the graph G. The cardinality of metric basis is called the metric dimension of G. In this paper, we develop algorithms to find the metric dimension and a metric basis of a simple graph. These algorithms have the worst-case complexity of O(nm). The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021. -
A Document Clustering Approach Using Shared Nearest Neighbour Affinity, TF-IDF and Angular Similarity
Quantum of data is increasing in an exponential order. Clustering is a major task in many text mining applications. Organizing text documents automatically, extracting topics from documents, retrieval of information and information filtering are considered as the applications of clustering. This task reveals identical patterns from a collection of documents. Understanding of the documents, representation of them and categorization of documents require various techniques. Text clustering process requires both natural language processing and machine learning techniques. An unsupervised spatial pattern identification approach is proposed for text data. A new algorithm for finding coherent patterns from a huge collection of text data is proposed, which is based on the shared nearest neighbour. The implementation followed by validation confirms that the proposed algorithm can cluster the text data for the identification of coherent patterns. The results are visualized using a graph. The results show the methodology works well for different text datasets. 2021, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.