An enhancement of cloud based sentiment analysis using svm based lexicon dictionary and adaptive resource scheduling

Abstract

Society is increasingly producing vast amount of data that may be either structured or unstructured. In recent years there has been a high demand for storing data in domains like finance, education, science and technology. Handling of huge and unstructured data (Big Data) may be a challenge that is a key to competitive advantage. Cloud computing and big data is considered as fastest moving technologies that are interlinked with one another to provide a solution for decision making problems. With the invention of new data mining techniques and machine learning algorithms analyzing big data is very critical. Big Data analysis is however a more challenging process than locating, preprocessing, identifying, understanding, and citing data. This thesis fills the research Gap observed towards a classification of a collection of unstructured or semi structured information with dynamic schema. Deployment on cloud computing infrastructures and indexing of information on the deep web is addressed. Visualization tool for cloud to handle large scale and complex data is addressed. Sentiment classification based on cloud customer feedback is developed by using the SVM with Lexicon based dictionary. newline