Analyzing online social networking to identify emotions and other characteristics
A comprehensive review of the various approaches to social networking user behavior analysis is reported in the International Journal of Computer Applications in Technology by a team from India. Pramod Bide and Sudhir Dhage of the Computer Engineering Department at the Sardar Patel Institute of Technology in Mumbai, Maharashtra, explain that various approaches can be used for gender prediction, the identification of malicious users, real-time user preference determination, and emotion detection.
Online social networks have been with us for many years now and they each have their pros and cons for their billions of users. For researchers and other observers they are a major source of data that can be used to glean insights into human behavior and the interactions and responses of users to others and to commercial, political, and other concerns that hope to engage with those users. As such, numerous analytical approaches have been tested that might extract insights from the various online social networks, each with its own successes and failures.
The team explains that hybrid techniques and ones that can be used to analyze behavior between networks can be the most powerful tools. The results that such approaches are able to glean about the networks' users can be useful to marketing departments, political campaigners, advocacy groups, and many other so-called "stakeholders" looking to make the most of the online world to fulfill their own agendas.
The team concedes that the vast majority of analytical tools focus on text-based updates on social networks, but some can take images and videos into consideration too, and even audio in some instances. Indeed, they suggest that the next step will be to survey tools that focus specifically on audio-visual content.
More information: Pramod Bide et al, Comprehensive survey of user behaviour analysis on social networking sites, International Journal of Computer Applications in Technology (2021). DOI: 10.1504/IJCAT.2021.119601