UNDERSTANDING ONLINE USER BEHAVIOR THROUGH MACHINE LEARNING AND INFORMATION-SEEKING ANALYTICS

Abstract

In the digital era, the analysis of online user behavior has become a critical aspect of understanding information consumption, personalization, and cybersecurity. This paper presents a machine learning–driven framework designed to classify online users based on their information-seeking behavior across digital platforms. The proposed approach integrates behavioral analytics, query intent modeling, and interaction features to uncover latent user patterns. Supervised and unsupervised learning algorithms, including Support Vector Machines (SVM), Random Forests, and K-Means clustering, are employed to identify distinctive behavioral classes and predict user intent with high accuracy. The model further leverages natural language processing (NLP) techniques to extract semantic cues from user queries and browsing content, thereby enhancing interpretability and contextual understanding. Experimental evaluations on benchmark datasets demonstrate that incorporating information-seeking attributes significantly improves the precision and recall of user classification compared to baseline methods. The findings suggest that machine learning combined with information-seeking analytics can provide valuable insights for user profiling, recommendation systems, and anomaly detection in online environments.