Patterns in the Growth and Thematic Evolution of Artificial Intelligence Research: A Study Using Bradford Distribution of Productivity and Path Analysis

Table 8

Yearwise listing of major application areas, their themes and different AI techniques used by researchers based on the available literature (2019).

Year	Area(s) of application	Theme(s)	Method(S)/models studied/used	Nature of the study	Inference/conclusion
2019	Disaster management	Spatial modelling of landslide susceptibility	Step-wise analysis ratio Analysis-ANFIS-	Comparative	SWARA-ANFIS-PSO more effective
			Shuffled frog Leaping algorithm or
			SWARA-ANFIS-SFLA
			SWARA-ANFIS-PSO
		Spatial prediction of wildfire probability	ANFIS-GA	Comparative	ANFIS-ICA more effective
			ANFIS-PSO
			ANFIS-SFLA
			ANFIS-imperialist competitive algorithm or ANFIS-ICA
		Flood susceptibility assessment	ANFIS: BAT algorithm or ANFIS-BA	Comparative	ANFIS-BBO more effective
			ANFIS-biogeography-based optimization or ANFIS-BBO
	Decision support/business analytics	Implications/roles of AI in business/management		Critical (philosophical) analysis	Highlighted internal and external implications of AI
				Critical (philosophical) analysis	Highlighted the role of AI personalized engagement marketing
				Critical (philosophical) analysis	Highlighted the implications of AI B2B marketing
	Water resource management	Saturated hydraulic conductivity in streambeds	ANN	Comparative	SVM more effective
			ANFIS
			SVM
	Environmental management (mining applications)	Blast-induced air over pressure prediction	Gradient boosting machine-RF-CA or GBM-RF-CA	Comparative	GBM-RF-CA more effective
			Empirical models
		Ground vibration prediction	Hierarchical K-means clustering (HKM)	Comparative	HKM-CA more effective
			Cubist algorithm (CA)
			HKM-CA (hybrid)
			RF
			SVM
			Classification and regression tree (CART)
			ANN (2-6-8-1)	Comparative	ANN (2-6-8-1) more effective
			KNN
			SVM
			CART
			HKM-ANN	Comparative	HKM-ANN more effective
			FCM-ANN
			HKM-SVR
			Empirical models