Exploration and analysis of structured and unstructured data using data science for accurate outcomes

Abstract

ABSTRACT newlineThis Thesis describes the research work specifically in the field of computer science for data conversion and their respective analysis using data science. As most of the companies are dealing with data driven projects and they try to analyze all kind of data present in the web today. As data is the most important asset today. All the organizations are busy in utilizing datasets for better outcomes and accurate outputs. Scientists are also busy in finding a better storage and data conversion techniques for better results. The aim of this study is to provide better techniques for data conversion specifically between unstructured data and structured data. Structured data is something which is among us from traditional data base and handled by many relational database systems. Whereas systems that support unstructured data are less and few of the non-relational database concepts now gave us a hope for data storage with different data structures. newlineFrom an extensive study of the literature, at first stage we tried to develop a technique that would convert an audible (unstructured data) into structured one using data science programming with various techniques in between. After converting and achieving structured data, at second stage machine learning algorithms are used that would try to analyze structured data such and their respective outputs helped to select the better algorithm among two. These algorithms are implemented with the help of data science language and their visualizations are used to reflect their differences. newlineThe research is basically done to overcome problem of data conversion and their analysis. It also used NOSQL database like MongoDB for the storage. This shows the significance of non-relational database into relational database system. The results confirm that we newlinecan convert an unstructured data(speech) into structured one using few basic method of computer science and can store our data directly into the non-relational database consisting of better scalability and performance. newlineResearch also reflects on the working of machine learning for data analysis where prediction of results is made on the past trained data. Here concept of data science, machine learning is used for data analysis and data conversion. newline