Showing papers in "International Journal of Intelligent Systems and Applications in 2017"
TL;DR: This model is based on sentiment analysis of financial news and historical stock market prices and provides better accuracy results than all previous studies by considering multiple types of news related to market and company with historical stock prices.
Abstract: Stock market prediction has become an attractive investigation topic due to its important role in economy and beneficial offers. There is an imminent need to uncover the stock market future behavior in order to avoid investment risks. The large amount of data generated by the stock market is considered a treasure of knowledge for investors. This study aims at constructing an effective model to predict stock market future trends with small error ratio and improve the accuracy of prediction. This prediction model is based on sentiment analysis of financial news and historical stock market prices. This model provides better accuracy results than all previous studies by considering multiple types of news related to market and company with historical stock prices. A dataset containing stock prices from three companies is used. The first step is to analyze news sentiment to get the text polarity using naïve Bayes algorithm. This step achieved prediction accuracy results ranging from 72.73% to 86.21%. The second step combines news polarities and historical stock prices together to predict future stock prices. This improved the prediction accuracy up to 89.80%.
105 citations
TL;DR: It is shown that the increase of efficiency is possible by adaptation of structure of the multilayer perceptron to the problem specification set and it is revealed that the structure adaptation lies in the determination the following parameters: 1. The number of hidden neuron layers; 2. the number of neurons within each layer.
Abstract: The paper is dedicated to the problem of efficiency increasing in case of applying multilayer perceptron in context of parameters estimation for technical systems. It is shown that the increase of efficiency is possible by adaptation of structure of the multilayer perceptron to the problem specification set. It is revealed that the structure adaptation lies in the determination the following parameters: 1. The number of hidden neuron layers; 2. The number of neurons within each layer. In terms of the paper, we introduce mathematical apparatus that allows conducting the structure adaptation for minimization of the relative error of the neuronetwork model generalization. A numerical experiment to demonstrate efficiency of the mathematical apparatus was developed and described in terms of the article. Further research in this sphere lies in the development of a method for calculation of optimum relationship between the number of the hidden neuron layers and the number of hidden neurons within each layer.
51 citations
TL;DR: The MENFN’s adaptive learning algorithm allows solving classification problems in a real-time fashion and its computational plainness in comparison with neuro-fuzzy systems and neural networks makes it effectual for solving the image recognition problems.
Abstract: An article introduces a modified architecture of the neo-fuzzy neuron, also known as a \"multidimensional extended neo-fuzzy neuron\" (MENFN), for the face recognition problems. This architecture is marked by enhanced approximating capabilities. A characteristic property of the MENFN is also its computational plainness in comparison with neuro-fuzzy systems and neural networks. These qualities of the proposed system make it effectual for solving the image recognition problems. An introduced MENFN’s adaptive learning algorithm allows solving classification problems in a real-time fashion.
46 citations
TL;DR: The paper analyzes modern methods of modeling impacts on information systems to determine the most effective approaches and use them to optimize the parameters of security systems and develops special software to verify the proposed method.
Abstract: The paper analyzes modern methods of modeling impacts on information systems, which made it possible to determine the most effective approaches and use them to optimize the parameters of security systems. And also as a method to optimize data security, taking in the security settings account (number of security measures, the type of security subsystems, safety resources and total cost information) allows to determine the optimal behavior in the ―impact-security‖. Also developed special software that allowed to verify the proposed method.
34 citations
TL;DR: The results obtained in this study indicate that there is an increase in accuracy of batik image classification using the artificial neural network with the combination of texture and shape features inBatik image.
Abstract: Batik is a text ile with mot ifs of Indonesian culture which has been recognized by UNESCO as world cultural heritage. Bat ik has many motifs which are classified in various classes of batik. This study aims to combine the features of texture and the feature of shapes’ ornament in batik to classify images using artificial neural networks. The value of texture features of images in batik is extracted using a gray level co-occurrence matrices (GLCM) which include Angular Second Moment (ASM) / energy), contrast, correlation, and inverse different moment (IDM). The value of shape features is extracted using a binary morphological operation which includes compactness, eccentricity, rectangularity and solidity. At this phase of the training and testing, we compare the value of a classification accuracy of neural networks in each class in batik with their texture features, their shape, and the combination of texture and shape features. From the three features used in the classification of batik image with artificial neural networks, it was obtained that shape feature has the lowest accuracy rate of 80.95% and the combination of texture and shape features produces a greater value of accuracy by 90.48%. The results obtained in this study indicate that there is an increase in accuracy of batik image classification using the artificial neural network with the combination of texture and shape features in batik image.
32 citations
TL;DR: The results of using local binary patterns for solving flames detecting problem are presented and modifications to improve the quality of detector work are proposed.
Abstract: Techniques to detect the flame at an early stage are necessary in order to prevent the fire and minimize the damage. The flame detection technique based on the physical sensor has limited disadvantages in detecting the fire early. This paper presents the results of using local binary patterns for solving flames detecting problem and proposes modifications to improve the quality of detector work. Experimentally found that using support vector machines classifier with a kernel based on Gaussian radial basis functions shows the best results compared to other SVM cores or classifier k-nearest neighbors.
30 citations
TL;DR: An adaptive neural system which solves a problem of clustering data with missing values in an online mode with a permanent correction of restorable table elements and clusters’ centroids is proposed in this article.
Abstract: An adaptive neural system which solves a problem of clustering data with missing values in an online mode with a permanent correction of restorable table elements and clusters’ centroids is proposed in this article. The introduced neural system is characterized by both a high speed and a simple numerical implementation. It can process information in a real-time mode.
30 citations
TL;DR: This paper proposes a novel meta-heuristic optimization algorithm inspired by general grass plants fibrous root system, asexual reproduction, and plant development that explores the bounded solution domain globally and locally.
Abstract: This paper proposes a novel meta-heuristic optimization algorithm inspired by general grass plants fibrous root system, asexual reproduction, and plant development. Grasses search for water and minerals randomly by developing its location, length, primary root, regenerated secondary roots, and small branches of roots called hair roots. The proposed algorithm explore the bounded solution domain globally and locally. Globally using the best grasses survived by the last iteration, and the root system of the best grass obtained so far by the iteration process and locally uses the primary roots, regenerated secondary roots and hair roots of the best global grass. Each grass represents a global candidate solution, while regenerated secondary roots stand for the locally obtained solution. Secondary generated hair roots are equal to the problem dimensions. The performance of the proposed algorithm is tested using seven standard benchmark test functions, comparing it with other metaheuristic well-known and recently proposed algorithms.
26 citations
TL;DR: Proposed framework provides a generalized solution for student employability prediction by integrating clustering and classification techniques and clearly depicts model performance over various classification techniques.
Abstract: Data Mining is gaining immense popularity in the field of education due to its predictive capabilities. But, most of the prior effort in this area is only directed towards prediction of performance in academic results only. Nowadays, education has become employment oriented. Very little attempt is made to predict students’ employability. Precise prediction of students’ performance in campus placements at an early stage can identify students, who are at the risk of unemployment and proactive actions can be taken to improve their performance. Existing researches on students’ employability prediction are either based upon only one type of course or on single University/Institute; thus is not scalable from one context to another. With this necessity, the conception of a unified model of clustering and classification is proposed in this paper. With the notion of unification, data of professional courses namely Engineering and Masters in Computer Applications students are collected from various universities and institutions pan India. Data is large, multivariate, incomplete, heterogeneous and unbalanced in nature. To deal with such a data, a unified predictive model is built by integrating clustering and classification techniques. TwoLevel clustering (k-means kernel) with chi-square analysis is applied at the pre-processing stage for the automated selection of relevant attributes and then ensemble vote classification technique with a combination of four classifiers namely k-star, random tree, simple cart and the random forest is applied to predict students’ employability. Proposed framework provides a generalized solution for student employability prediction. Comparative results clearly depict model performance over various classification techniques. Also, when the proposed model is applied up to the level of the state, classification accuracy touches 96.78% and 0.937 kappa value.
26 citations
TL;DR: An intelligent grouping algorithm is developed which utilizes machine learning generated clusters to form heterogeneous groups and has the advantage of dynamically changing the group membership based on learners’ collaboration competence level.
Abstract: The current Learning Management Systems used in e-learning lack intelligent mechanisms which can be used by an instructor to group learners during an online group task based on the learners‟ collaboration competence level. In this paper, we discuss a novel approach for grouping students in an online learning group task based on individual learners‟ collaboration competence level. We demonstrate how it can be applied in a Learning Management System such as Moodle using forum data. To create the collaboration competence levels, two machine learning algorithms for clustering namely Skmeans and Expectation Maximization (EM) were applied to cluster data and generate clusters based on learner‟s collaboration competence. We develop an intelligent grouping algorithm which utilizes these machine learning generated clusters to form heterogeneous groups. These groups are automatically made available to the instructor who can proceed to assign them to group tasks. This approach has the advantage of dynamically changing the group membership based on learners‟ collaboration competence level.
24 citations
TL;DR: A new method of optimal ICN structuring is developed which unlike the existing ones involves the variant synthesis of structures hierarchy and formalizing selection problems on the basis of settheoretic models, which enables providing the efficiency of application of information and technical net resources.
Abstract: The subject matter of the article is developing information and communication network (ICN) for critical infrastructure systems (CIS). The aim of the work is to provide high-quality information and telecommunication processes by developing the optimal version of distributing CIS functional tasks and ICN processes to the network nodes. The article deals with following problems: developing a model for mapping the information and technical ICN structures, developing a method for variant synthesis of ITS structural models, a formalized representation of the problem of selecting CIS optimal structure. The methods used are: the system method, the set-theoretic and graphic analytic approaches, methods of hierarchic structures synthesis, optimization methods. The following results were obtained: the use of system approach for formalizing the information processing process in CIS was justified; mapping the ICS functional system into the information and technical one was presented as multilevel graph chain; the generalized representation of graph structures hierarchy was developed for the set of data transmitting tasks; this approach enabled formal representing alternative variants that consider the main links, sequencing, the amount and flows of the processed information among the different structure levels; the scheme of variant synthesis method of ICN models according to graph structures mapping was developed; the problem of selecting optimal ICN structures was formally presented; a complex efficiency criterion for solving problems of optimizing variant synthesis of structures; the problem of optimal synthesis of the structure of the given level factored in resource constraints was formulated. Conclusions. The article deals with such novelty aspects as improving the model of problem of selecting the optimal ICN structure by settheoretic formalization factored in the criterion of maximum intensity of computational resource application, which enabled determining structural links among the major elements considering the decomposition of the model up to the basic elements such as \"node\" and \"task\" and the development of a new method of optimal ICN structuring which unlike the existing ones involves the variant synthesis of structures hierarchy and formalizing selection problems on the basis of settheoretic models, which enables providing the efficiency of application of information and technical net resources.
TL;DR: Fuzzy clustering procedures for categorical data are proposed in the paper and a detailed description of a possibilistic fuzzy clustering method based on frequency-based cluster prototypes and dissimilarity measures for category data is given.
Abstract: Fuzzy clustering procedures for categorical data are proposed in the paper. Most of well-known conventional clustering methods face certain difficulties while processing this sort of data because a notion of similarity is missing in these data. A detailed description of a possibilistic fuzzy clustering method based on frequency-based cluster prototypes and dissimilarity measures for categorical data is given.
TL;DR: Evaluation of time series similarity is based on Dynamic Time Warping and provides various solutions for Content Based Video Information Retrieval.
Abstract: Temporal clustering (segmentation) for video streams has revolutionized the world of multimedia. Detected shots are principle units of consecutive sets of images for semantic structuring. Evaluation of time series similarity is based on Dynamic Time Warping and provides various solutions for Content Based Video Information Retrieval. Time series clustering in terms of the iterative Dynamic Time Warping and time series reduction are discussed in the paper.
TL;DR: The proposed methodology uses Radial Basis Function Network (RBFN) and Feedforward based Back propagation network (BPN) algorithm to generate pitch angle and is able to compensate the nonlinear characteristic of wind speed.
Abstract: This paper proposes an advanced pitch angle control strategy based on neural network (NN) for variable speed wind turbine. The proposed methodology uses Radial Basis Function Network (RBFN) and Feedforward based Back propagation network (BPN) algorithm to generate pitch angle. The performance of the proposed control technique is analyzed by comparing the results with Fuzzy Logic Control (FLC) and Proportional Integral (PI) control techniques. The control techniques implemented is able to compensate the nonlinear characteristic of wind speed. The wind turbine is smoothly controlled to maintain the generator power and the mechanical torque to the rated value without any fluctuation during rapid variation in wind speed. The effectiveness of the proposed control strategy is verified using MATLAB/Simulink for 2-MW permanent magnet synchronous generator (PMSG) based wind energy conversion system.
TL;DR: The method of multicriteria optimization is examined with determination of resulting objective function, which allows to consider various aspects of the problems of media choice and optimal budgeting and budget allocation simultaneously to get a satisfactory solution of the problem.
Abstract: A concept targeting information technology is considered. The model of selection of the optimal amount of advertising on various Internet resources, in order to maximize the desired reach to the target audience is analyzed which. This model is different from traditional. A chance constrained target programming model was developed after considering the parameter that corresponds to reach for different media as random variables. The random variables in this case has been considered as the values with known mean and standard deviation. The reachability parameter can be determined by finding the ideal solution and the law on which the parameter values change. The method of multicriteria optimization is examined with determination of resulting objective function, which allows to consider various aspects of the problems of media choice and optimal budgeting and budget allocation simultaneously to get a satisfactory solution of the problem.
TL;DR: The development of a decentralized mechanism for the resources control in a distributed computer system based on a network-centric approach and performs the analysis of its operation parameters in compare with resource control modules implemented on the hierarchical approach and on the decentralized approach with the creation of the communities of the computing resources.
Abstract: In this paper, we present the development of a decentralized mechanism for the resources control in a distributed computer system based on a network-centric approach. Intially, the network-centric approach was proposed for the military purposes, and now its principles are successfully introduced in the other applications of the complex systems control. Due to the features of control systems based on the network-centric approach, namely adding the horizontal links between components of the same level, adding the general knowledge control in the system, etc., there are new properties and characteristics. The concept of implementing of resource control module for a distributed computer system based on a network-centric approach is proposed in this study. We, basing on this concept, realized the resource control module and perform the analysis of its operation parameters in compare with resource control modules implemented on the hierarchical approach and on the decentralized approach with the creation of the communities of the computing resources. The experiments showed the advantages of the proposed mechanism for resources control in compare with the control mechanisms based on the hierarchical and decentralized approaches.
TL;DR: This paper has given a review about how the reversible logic and QCA nanotechnology together result in ultra-low power designs and optimized the design of Peres reversible gate using the concepts of explicit interaction of cells in QCA and verified the universal functionality using the optimized designs.
Abstract: The essence of the technology business lies in the improvements and advancements that are continuously taking place in the industry. From vacuum tubes, diodes and transistors to the concepts of nano level designing have by and large created a revolution in the history of mankind. The biggest milestone in this journey has been the CMOS technology which has managed to survive for decades and is still an ongoing research area. However, advancing the technology includes many other dimensions which need to be taken care of. As the devices go on decreasing in size with the improving technology the power dissipation in them becomes a major issue. To counter this, a new logic called reversible logic has come into the pool of research. Further a shift from the transistor based paradigm is being explored to go down to ultra-small structures. A major breakthrough in this can be the Quantum Dot Cellular Automata (QCA) Nanotechnology. In this paper we have given a review about how the reversible logic and QCA nanotechnology together result in ultra-low power designs. Further we have optimized the design of Peres reversible gate using the concepts of explicit interaction of cells in QCA and verified the universal functionality using the optimized designs.
TL;DR: This paper presents many challenges and issues observed during sentiment analysis of On-line Microtexts by using the established methods of sentimentAnalysis of unstructured reviews.
Abstract: With the evolution of World Wide Web (WWW) 2.0 and the emergence of many micro-blogging and social networking sites like Twitter, the internet has become a massive source of short textual messages called on-line micro-texts, which are limited to a few number of characters (e.g. 140 characters on Twitter). These on-line micro-texts are considered as real-time text streams. Online micro-texts are extremely subjective; they contain opinions about various events, social issues, personalities, and products. However, despite being so voluminous in quantity, the qualitative nature of these micro-texts is very inconsistent. These qualitative inconsistencies of raw on-line micro-texts impose many challenges in sentiment analysis of on-line micro-texts by using the established methods of sentiment analysis of unstructured reviews. This paper presents many challenges and issues observed during sentiment analysis of On-line Microtexts.
TL;DR: A heuristic algorithm for automatic navigation and information extraction from journal’s home page has been devised and the experimental results show that the heuristic technique provides promising results with respect to precision and recall values.
Abstract: WWW is a huge repository of information and the amount of information available on the web is growing day by day in an exponential manner. End users make use of search engines like Google, Yahoo, and Bingo etc. for retrieving information. Search engines use web crawlers or spiders which crawl through a sequence of web pages in order to locate the relevant pages and provide a set of links ordered by relevancy. Those indexed web pages are part of surface web. Getting data from deep web requires form submission and is not performed by search engines. Data analytics and data mining applications depend on data from deep web pages and automatic extraction of data from deep web is cumbersome due to diverse structure of web pages. In the proposed work, a heuristic algorithm for automatic navigation and information extraction from journal’s home page has been devised. The algorithm is applied to many publishers website such as Nature, Elsevier, BMJ, Wiley etc. and the experimental results show that the heuristic technique provides promising results with respect to precision and recall values.
TL;DR: The proposed deep hybrid system of computational intelligence with architecture adaptation for medical fuzzy diagnostics allows to increase a quality of medical information processing under the condition of overlapping classes due to special adaptive architecture and training algorithms.
Abstract: In the paper the deep hybrid system of computational intelligence with architecture adaptation for medical fuzzy diagnostics is proposed. This system allows to increase a quality of medical information processing under the condition of overlapping classes due to special adaptive architecture and training algorithms. The deep hybrid system under consideration can tune its architecture in situation when number of features and diagnoses can be variable. The special algorithms for its training are developed and optimized for situation of different system architectures without retraining of synaptic weights that have been tuned at previous steps. The proposed system was used for processing of three medical data sets (dermatology dataset, Pima Indians diabetes dataset and Parkinson disease dataset) under the condition of fixed number of features and diagnoses and in situation of its increasing. A number of conducted experiments have shown high quality of medical diagnostic process and confirmed the efficiency of the deep hybrid system of computational intelligence with architecture adaptation for medical fuzzy diagnostics.
TL;DR: Simulation results shows that the proposed algorithm with particle swarm optimization approach outperform the other existing algorithms of its category and shows their supremacy in term of alive nodes, energy expenditure, packet delivery ratio, and throughput of network.
Abstract: In the previous years, wireless sensor networks (WSNs) got lot of attraction from the scientific and industrial society. WSNs are composed of huge number of small resource constrained devices recognized as sensors. Energy is a vital issue in WSN. Energy efficient clustering is an eminent optimization problem which has been studied extensively to prolong the lifetime of the network. This paper demonstrates the programming formulation of this problem followed by a proposed algorithm with particle swarm optimization (PSO) approach. The clustering method is stated by taking into consideration of energy saving of nodes. The proposed algorithm is experimented widely and results are evaluated with existing methods to show their supremacy in term of alive nodes, energy expenditure, packet delivery ratio, and throughput of network. Simulation results shows that our proposed algorithm outperform the other existing algorithms of its category.
TL;DR: In this article, an approach based on memberhsip and likelihood functions sharing has been proposed for ordinal scale data clustering under conditions of overlapping clusters, which is characterized by robustness to outliers due to a way of ordering values while constructing membership functions.
Abstract: A task of clustering data given in the ordinal scale under conditions of overlapping clusters has been considered. It's proposed to use an approach based on memberhsip and likelihood functions sharing. A number of performed experiments proved effectiveness of the proposed method. The proposed method is characterized by robustness to outliers due to a way of ordering values while constructing membership functions.
TL;DR: A hybrid bee colony algorithm is proposed in this paper for generating and optimizing the test cases from combinational UML diagrams.
Abstract: To detect faults or errors for designing the quality software, software testing tool is used. Testing manually is an expensive and time taking process. To overcome this problem automated testing is used. Test case generation is a vital concept used in software testing which can be derived from requirements specification. Automation of test cases is a method where it can generate the test cases and test data automatically by using search based optimization technique. Model-driven testing is an approach that represents the behavioral model and also encodes the system behavior with certain conditions. Generally, the model consists of a set of objects that defined through variables and object relationships. This piece of work is used to generate the automated optimized test cases or test data with the possible test paths from combinational system graph. A hybrid bee colony algorithm is proposed in this paper for generating and optimizing the test cases from combinational UML diagrams.
TL;DR: In this paper, a fuzzy clustering algorithm for multidimensional data is proposed in which the data is described by vectors whose components are linguistic variables defined in an ordinal scale.
Abstract: A fuzzy clustering algorithm for multidimensional data is proposed in this article. The data is described by vectors whose components are linguistic variables defined in an ordinal scale. The obtained results confirm the efficiency of the proposed approach.
TL;DR: The proposed method of graphical data protection is a combined crypto-steganographic method based on a bit values transformation according to both a certain Boolean function and a specific scheme of correspondence between MSB and LSB.
Abstract: The proposed method of graphical data protection is a combined crypto-steganographic method. It is based on a bit values transformation according to both a certain Boolean function and a specific scheme of correspondence between MSB and LSB. The scheme of correspondence is considered as a secret key. The proposed method should be used for protection of large amounts of secret graphical data.
TL;DR: The method of multicriterion optimization of multilevel networks is examined with determination of resulting objective function, which allows to carry out the synthesis of control system software-defined network (SDN) in the conditions of unforeseen changes of structure of the system.
Abstract: A concept software-defined network is considered. Architecture of software-defined network is analyzed which, differently from traditional, foresee the separation of C-plane from a plane communication of data. The method of multicriterion optimization of multilevel networks is examined with determination of resulting objective function, which allows to carry out the synthesis of control system software-defined network (SDN) in the conditions of unforeseen changes of structure of the system.
TL;DR: Two important types algorithm are compared to identifying the highly nonlinear systems, namely: Auto-Regressive with eXternal model input(ARX) and Auto Regressive moving Average with external modelinput (Armax) Theory.
Abstract: System Identification is used to build mathematical models of a dynamic system based on measured data. To design the best controllers for linear or nonlinear systems, mathematical modeling is the main challenge. To solve this challenge conventional and intelligent identification are recommended. System identification is divided into different algorithms. In this research, two important types algorithm are compared to identifying the highly nonlinear systems, namely: Auto-Regressive with eXternal model input(ARX) and Auto Regressive moving Average with eXternal model input (Armax) Theory. These two methods are applied to the highly nonlinear industrial motor.
TL;DR: The presented survey though provides an insight towards the fundamentals of big data analytics but aims towards an analysis of various optimization techniques used in map reduce framework and big data Analytics.
Abstract: Data is one of the most important and vital aspect of different activities in today's world. Therefore vast amount of data is generated in each and every second. A rapid growth of data in recent time in different domains required an intelligent data analysis tool that would be helpful to satisfy the need to analysis a huge amount of data. Map Reduce framework is basically designed to process large amount of data and to support effective decision making. It consists of two important tasks named as map and reduce. Optimization is the act of achieving the best possible result under given circumstances. The goal of the map reduce optimization is to minimize the execution time and to maximize the performance of the system. This survey paper discusses a comparison between different optimization techniques used in Map Reduce framework and in big data analytics. Various sources of big data generation have been summarized based on various applications of big data.The wide range of application domains for big data analytics is because of its adaptable characteristics like volume, velocity, variety, veracity and value .The mentioned characteristics of big data are because of inclusion of structured, semi structured, unstructured data for which new set of tools like NOSQL, MAPREDUCE, HADOOP etc are required. The presented survey though provides an insight towards the fundamentals of big data analytics but aims towards an analysis of various optimization techniques used in map reduce framework and big data analytics.
TL;DR: A novel data collection technique for prediction of election outcomes and a topic modeling method for extracting topics and an intelligent prediction model based primarily on the volume of tweets and sentiment of users are presented.
Abstract: Exploiting social media data by extracting key information from it is one of the great challenges in data mining and knowledge discovery. Every election campaign has an online presence of voters which uses these social media platform to express their sentiments and opinions towards political parties, leaders and important topics. This paper presents a novel data collection technique for prediction of election outcomes and a topic modeling method for extracting topics. Data collection technique used RSS (Rich Site Summary) feeds of news articles and trending keywords from Twitter simultaneously and constructed an intelligent prediction model based primarily on the volume of tweets and sentiment of users. This paper effort to improve electoral predictions using social media data based dynamic keyword methodology. Different techniques for electoral prediction based on social media data has been investigated based on existing literature and isolate the factors which improve our methodology. Meaningful inferences such as the popularity of leaders and parties during different intervals, trending issues, and important factors are extracted from the data set. The election outcomes are compared with traditional methods used by survey agencies for exit polls and validation of results showed that social media data can predict with better accuracy. The research has identified that data collection technique and timing play an important role in yielding better accuracy in predicting outcomes and extracting meaningful inferences.