An Effective System for Resume Analysis and Personality Prediction Using SVM

The human persona has performed a vital function in a character's existence and in the growth of an industry. One method to decide about the human character is the use of preferred questionnaires or via studying the resumes. Many algorithms have been utilized up to this point to extract words from unprocessed text input for use in personality prediction systems; some of them make use of composite models and deep learning methods. The following are some drawbacks of machine learning (ML) algorithms like random forest and logistic regression, notwithstanding their usefulness. Due to sequential input, these techniques first require a lot of time to train the model. Second, because these algorithms are unable to accurately express the actual significance of words so, some context of the words is lost. In this paper, we gift a new system that analyzes the resume of the candidate and predicts his or her personality. The system uses the submitted resume to predict the professional personality. The preprocessing is done using BERT and XLNet which are pre-trained language frameworks. Then extracted features are given to the trained support vector machine (SVM) model. SVM is selected as it provides the best accuracy. The system will output a predicted personality for the resume having 92% accuracy.