DEEP LEARNING APPROACH FOR PREDICTING AD CLICK

Abstract

In the digital advertising landscape, predicting user behaviour and user interest impact in response and action based on the advertisement. This study used a deep learning approach for predicting ad clicks, leveraging a comprehensive dataset that includes user demographics, ad attributes, and contextual information. To date, existing methods for Ad click prediction, or click-through rate prediction, mainly consider representing users as a static feature set and train machine learning classifiers to predict clicks. In this paper we propose a deep learning approach for predicting ad click by BI-LSTM method and evaluate the model performance metrics such as precision, recall and f1-score.our goal is to accurately predict the past and future behaviour of the user based on the ads. To achieve the goal, we collect page information displayed to the users as a temporal sequence and use bi direction long short-term memory (BI-LSTM) network to learn features of both forward and backward direction that represents user interests as latent features. on real-world data show that, compared to existing approaches, considering bidirectional long and short-term sequences, user requests results in improvements in user Ad response prediction and campaign specific user Ad click prediction.