Novel Approaches To Mine Facebook Data For Effective Digital Marketing Strategies

Abstract

newline Social Media is a huge manifesto and an interface for number of social applications alike Facebook, Twitter, LinkedIn and so on. These applications have gained tremendous success in generating information by providing users a flexible tool to share one s ideas, feelings, thoughts, or the opinion a user wants to share with others. Every social media application is having its own kind of interface, structure, formation and storage. Among all social applications, top ten leading applications as of 2021 are: Facebook, Instagram, Twitter, LinkedIn, Printerest, YouTube, Snapchat, TikTok, Reddit and WhatsApp. Users of any of these social media platforms are billions in count, Facebook alone stores four petabytes of data every day which means million gigabytes and the data is stored in Hive that can carry around 300 petabytes of data. Per user, Facebook has provided 500 megabytes of storage space and the data space may increase in further days. Likewise, all other social media applications have been assigned some sort of storage for the users. The uploaded data does not contain proper structure or it may not be in order and contains huge amount of noise, such data cannot be processed further hence, extracting the useful data or knowledge from such vast data is an essential work for data scientists. Using the discovered knowledge, several helpful predictions can be made and these extracted patterns give essential information regarding every field such as marketing, education, health science, business, banking, industry, corporation, finance so on as social media has become a flexible and user-friendly platform for online business advertisements. The entire work and all the experiments of present research explain how social media applications are playing vital role in advertising different aspects or objects in various fields. Several experiments have been conducted in the current work on Facebook dataset as Facebook stores petabytes of data which leads to gather huge volume of noise and faces most difficulti