Collaborative Filtering Based-Recommender System Using Ant Colony Optimisation for Path Planning

In recent years, the recommender system has been extraordinarily successful in the Internet industry, and it has benefited many webs application developers. Popular websites such as YouTube, Netflix, and Facebook use the recommender system to enhance the user experience and attract new users. Nevertheless, the demand for pervasive information processing over the Web has paved the way for a context-aware recommender system capable of dealing with a massive amount of data and capable of generating an effective and personalised output. In this paper, a recommender system was designed and developed for the tourism industry in New Zealand. The proposed method comprises collaborative filtering techniques and a machine learning model, specifically the k-nearest neighbours (k-NN) algorithm, as the framework of the similarity calculation to analyse user preference data and predict similar tourist attractions. In addition, the Ant Colony optimisation algorithm (ACO) was utilised to aggregate the shortest path between multiple attraction points. Furthermore, the researchers demonstrate the integration process of the recommender system into the website and travel route module. This research is based on Python and JavaScript and uses Flask, NumPy and Vue.js to build the web-based recommender system.