Improved Recommender System for Kid's Hobby Prediction using different Machine Learning Techniques

A recommendation system plays a vital role in the social media world, and it is used in many applications which provide recommendations based on user preferences. A recommendation is nothing but a choice-making system. The utilization of recommender systems is increasing daily as many choice-based challenges potentially influence users. Now a day's, parents are worried about their kids regarding the future. Most parents need help to identify their kid's interests and force their interests against kids. So this prediction will be helpful to society. There are many types of recommendation systems with different applications that have been successfully adopted. This work needs to specify the age limits for kids, and it's preferable for below ten years old kids. The dataset is prepared based on different parameters. Fourteen parameters have been identified, and the result is three possible classifications. The first classification is 1) academics, 2) arts, and 3) sports. The academics indicate the kid's interest in science with a mathematics background. A total of 1600 dataset has been used for training and modelling the different algorithm. Azure and auto-learn tools are used for processing the data, and n number of classification algorithms are used. In azure Multiclass Logistic Regression produced the highest result of 93.21%. In auto-sklearn algorithms, Liblinear SVC (support vector classification) produced 93% of the result.