Age-group, gender, handedness and number of hands used are the common personality traits of a typist, and identifying such traits can be a key in identifying the person in today’s fast world. This particular piece of work is the objective, i.e., an indicative pathway, toward that goal by monitoring and analyzing the way a user types on a touch screen of a smartphone. Study of such traits and analyzing the typing pattern on a conventional computer keyboard has been investigated well. But the conventional keyboard is being replaced with the advent of smartphones with a variety of features, low cost and portability. Therefore, identifying traits through the touch screen is more significant and might be notably beneficial for personal identity prediction and verification. In this paper, we discuss the data acquisition method, classification approach and the evaluation process which are found as more appropriate to discover the trait identities to be used in variety of Web-based applications specifically in the area of e-commerce, online examination, digital forensics, targeted advertisement, age-restricted access control, human–machine interaction, social networks, user identity verification akin to biometrics. Multiple machine learning (ML) methods were used to develop the model, and more suitable and practical evaluation test option—leave-one-user-out cross-validation—was used to check the validity of the proposed model. The efficacy of our approach is illustrated on the dataset collected in the Web-based environment from 92 volunteers. The probability of predicting a user with such traits has also been illustrated here. The study shows timing features of primary keystroke dynamics incorporated with the traits, and the user identification accuracy can be gained up to 17%. © 2018, Springer-Verlag London Ltd., part of Springer Nature.