Self Organizing Map (SOM) is an artificial and unsupervised neural network that refers to the process of representing a high-dimensional data set as a two-dimensional discretized pattern. The main goal of Self Organizing Map (SOM) is to reduce dimensionality while preserving the topological structure of data in the original feature space. Merits of SOM are easy in data interpretation and similarity observation of data. It is the type of competitive learning network and one of the classification techniques as it groups or clusters data points based on the similarity of the data points, without knowing the class memberships of the input data. SOM utilizes clustering and mapping techniques to map the high dimensional data onto two-dimensional data to interpret the complex problem easily. SOM architecture comprises input and output layers with fully connected neurons and uses a competitive learning algorithm to distribute input samples. The SOM implements an ordered dimensionality-reducing map of the data and follows the probability density function of the data. Widely used applications of SOM are data visualization and data pattern classification. The application areas of Self Organizing map are pattern recognition, image analysis, process monitoring and control, anomaly detection, medical diagnosis, virus attack detection, and fault diagnosis. Future development approaches of SOM are intrusion detection system with accurate topological data representation, Growing Hierarchical Self-Organizing Map (GHSOM) for hierarchical analysis, analysis of spatial and temporal aspects based spread of coronavirus using SOM, SOM text clustering on large datasets, among others.