Automation of detection of social network mental disorders – A review

Abstract

Recent years have seen a growing association between social media use and mental disorders. Experts opine that excessive use of social media sites has reached alarming levels, leading to Social Network Mental Disorders (SNMD). The Social Networking Sites (SNS) usage generates massive amount of complex data, which is difficult to analyze, find patterns within and make predictions manually, in turn making it difficult to detect SNMD in the SNS users. The paper summarizes the studies involved in the detection of mental disorders due to prolonged use of social networking sites and provides means of understanding an accurate predictive platform using data mining techniques to build machine-learning framework complementary to the conventional detection methods. We present an in-depth survey of the proposals that included the subtypes of SNMD in the SNS users – Cyber relationship addiction, Net compulsion, Information overload addiction, Cyber sex addiction and Computer addiction. The researchers bring forth models that integrate data mining techniques, Natural Language and computer vision programming tools, with social media and behaviour sciences, providing promising results. Data mining approaches to detect SNMD despite being challenging, are effective and can be used as an efficient tool. The challenges and issues related to automate the detection process is also analyzed in the paper. Thus we present that the automation of detection of SNMD in SNS users has potential to improve the existing health care systems.