Data Analytics in Mental Healthcare

Table 3

Data analytics and predicting mental health.

Authors	Tool/technology	Methodology	Purpose of finding	Strength	Weakness
Dimitrov [5]	(i) Sensing technology	Emergence of medical internet of things (mIoT) in existing mobile apps	Providing benefits to the customers	(i) Achieving improved mental health	Adding up garbage data to the sensors
	(ii) Artificial intelligence		(i) Avoiding chronic and diet-related illness
			(ii) improving cognitive function	(ii) improving lifestyles in real time decision-making
Monteith et al. [6]	Survey based approach	Clinical data mining	Analyzing different data sources to get psychiatry data	Optimized precedence opportunities for psychiatry	N/A
Kellmeyer [7]	Neurotechnology	(i) Machine learning	Enhancing the security of devices and sheltering the privacy of personal brain	Maximizing medical knowledge	Model needs huge amount of training data as brain disease is rarely captured
		(ii) Consumer-directed neurotechnological devices
		(iii) Combining expert with a bottom-up process
Yang et al. [9]	Long-term monitoring wearable device with internet of things	Well-being questionnaires with a group of students	Developing app-based devices linked to android phones and servers for data visualization monitoring and environment sensing	Perfectly working on long-term data	Offline data transfer instead of real time
Monteith and Glenn [10]	Automated decision-making	Hybrid algorithm that combines the statistical focus and data mining	Tracking day-to-day behavior of the user by automatic decision-making	Automatically detecting human decision without any input	How to ignore irrelevant information is a key headache
De Beurs et al. [14]	Online intervention	Expert-driven method Intervention mapping Scrum	Increasing user involvement under limited resources	Standardizing the level of user involvement in the web-based healthcare system	Deciding threshold for user involvement is problematic
Kumar and Bala [12]	Hadoop	Doing sentimental analysis and saving data on Hadoop	Analyzing twitter users’ view on a particular business product	Checking out popularity of a particular service	Usage of two programming languages needs experts
Goyal [11]	KNN and Naïve Bayes classifier	Text mining and hybrid approach combining KNN and Naïve Bayes	Opinion mining of tweets related to food price crisis	Cost-effective way to predict prizes	Data needs to be cleaned before training