URL Classification based on Active Learning Approach
Rakotoasimbahoaka Antsa Cyprienna, Raharijaona Zo Lalaina Yannick, Iadaloharivola Randria, Razafindrakoto Nicolas Raft · 2021
Web address URL fractions are commonly used in building URL detection techniques. However, the problem of training efficiency is still unsolved despite different models proposed in the literature. As training machine learning models require large quantities of training data and labeling data is time-consuming besides being expensive, using active learning models can be a powerful tool to overcome this situation. This paper investigated the efficiency of active learning using query-based committee strategy. An experiment was carried out; active learning outperforms state-of-the-art methods such as classical machine learning algorithms in terms of performance.