Abstract:
Information visualization is the final process of any data analytics project, which gives an overall insight about the identified trends, patterns and anomalies. Efforts in data
analysis become useless when we fail to convey the results effectively. Among visualization methods graphs play a vital role and are used frequently. This paper introduces “Roopana”, a semi-automated platform for data visualization. It is capable of
identifying the context of user data automatically, consider the purpose of data visualization (comparison, distribution, relationship, composition) and recommend the most accurate chart type to use. Considering chart types, “Roopana” is capable
of proving recommendations from large number of charts. It then enables the use of less frequently used chart types, which are matching with data context and purpose of visualization. Further the system can alter the order of recommendations depending on
user feedback enabling the most popular chart types to appear on top. Thus data scientists and developers can use “Roopana” to select most appropriate chart for their data without having an expert knowledge in the field of data visualization projects they
are working on.
