The point of interest (POI) recommendation for mobile digital culture heritage (M-DCH) based on the behavior analysis using the recurrent neural networks (RNN) and user-collaborative filtering

Many Point Of Interest (POI) recommendation systems need to collect past users’ scoring to generate the recommendation for current users. It always results in the not so precise recommendation because not a lot of users are willing to do the scoring. With the advanced deep learning technique, this work proposes the POIs’ recommendation method that doesn’t require a scoring mechanism to have the great precision, recall and diversity. The proposed POIs’ recommendation method utilizes the deep learning model to analyze user’s operational behaviors and then judge the user’s preference. As a result, the proposed POI’s recommendation method (i) can be built in an environment without a scoring mechanism because it can catch the user’s preferences by analyzing his operational behaviors and (ii) considers similar users’ historical data to make the recommended results more diversity. The performance evaluation shown that the precision, recall, f1-score and the next POI predicted rate of the proposed method is better than that of the Multi-Layer Perceptrons (MLPs) and the Long Short-Term Memory (LSTM) models. The diversities of the proposed method’s results are better than that of the LSTM model. Therefore, the proposed method balances the precision, recall and diversities.