基于互联网信息的多约束多目标旅游路线推荐

针对新游客在陌生城市如何规划旅游路线的问题,研究基于景点评分机制以及用户多约束的旅游路线推荐问题。首先提取景点的开放时间、门票与GPS坐标等及旅游网站上对于景点的评价信息等;然后提出一种基于多约束的k贪心算法,可以为游客推荐较好的旅游线路,并有效消除了推荐系统对先验知识的依赖。以驴评网上北京著名景点的信息作为数据集,实现并评估了推荐算法。实验结果表明,该方法能够为用户提供准确合理的路线规划。

Multi-constraint and multi-objective trip recommendation based on internet information

To help a novice visitor to make a travel route plan in an unfamiliar city, we study on how to evaluate a prospective scenery spot and recommend visitors with multiple constraints and diverse objectives. We first extract specific information of an attraction from the Tourism website, including the score graded by visitors, opening time, the price of an entrance ticket, and the GPS coordinates. Afterwards, we propose a k greedy algorithm to generate a feasible trip recommendation with good performance, i.e. low cost and long stay time. Based on the datasets of the famous attractions in Beijing collected from the Lvping website, we implement and evaluate the proposed algorithm. Experimental results show that it can provide accurate and reasonable trip plans for users with diverse requirements.