Extracting patterns from un structured data using Data mining algorithms and sentiment analysis

Abstract

Text mining confronts difficulties due to the information s extremely unstructured newlineformat and challenge of statistically evaluating a large collection of documents, such as big newlinedata. A popular area of research in the field of text mining is Sentiment Analysis (SA). It newlineuses advanced computational approach to address opinions, sentiments and text s newlinesubjectivity. Currently, there are technologies to extract, analyze and keep track of the newlinesubjective opinions on text and document data, as well as to evaluate and track sentiment or newlineemotions. Data pre-processing based on Sentiment Analysis (SA) is carried out on tweets in newlineorder to build Linear Support Vector Machines (LSVM), Naïve Bayes (NB) classifiers and newlineassess the view of a textual data. According to experimental findings, linear support vector newlinemachines generate better outcomes over naïve bayes classifiers. Linear support vector newlinemachine classifier has been tested, in particular on Twitter demonetization reviews. newlineThe sentiment analysis task and the polarity of the reviews - positive, negative and neutral newlinehas been obtained. The outcomes of another model i.e. Hierarchal Clustering (HC) are newlinecontrasted with Naïve Bayes Classification (NBC) algorithms. In comparison to naive bayes newlineclassifiers, hierarchical clustering has proved to produce more qualitative results. The test newlinedataset has assessed this based on the typical proportion of user reviews that indicate their newlinethoughts on a particular location. There are a number of real-world data sets, such as newlinemoviepang02 and chicagoaffnia that are utilized in machine learning methods for newlinecategorization and clustering. However, there is still a need for new research on the newlineincreasing use of social networking media and microblogging websites as data sources. On newlinecomparing the outputs of Naïve Bayes (NB) classifiers with those obtained by Multinomial newlineNaïve Bayesian (MNB) models across numerous Amazon product categories, naïve bayes newlineclassifiers generate more appropriate outcomes. Using the naïve bayes classification newlinetechnique, we analy