基于标签的个性化旅游推荐及情感分析

Personalized Tourism Recommendation and Emotion Analysis Based on Tags

随着旅游业的快速发展,旅游信息也变得多样化,在当今的快节奏时代,更多的游客希望能够快速找到最优的方案。为了解决这一问题,从海量的数据信息中筛选出有用的数据,根据各个景点的简介提取出景点特色打上标签,以便更快速地查找相关信息,为个性化旅游推荐做铺垫。对用户进行情感分析不仅可以根据好感度的结果来判断是否继续推荐与历史相似的景点,还可以将大范围的信息进行精细化过滤,为用户提供一个更加精准的推荐,最后根据过滤信息进行相关景点推荐或者重新推荐热门景点。文章主要使用语言模型将标签词语转化为向量,然后训练向量,使用余弦相似度计算出相似度较高的标签,融入情感分析,从而完成相似景点的个性化推荐。

With the development of tourism,tourism information has become complex and diversified.In today’s fast-paced era,more tourists hope to find the optimal plan quickly and conveniently.In order to solve this problem,useful data are screened out from the massive data information,and the characteristics of scenic spots are extracted and labeled according to the brief introduction of each scenic spot,so as to facilitate faster search and screening,paving the way for personalized tourism recommendation.The user’s sentiment analysis can not only judge whether to continue to recommend scenic spots similar to the history based on the results of the goodwill,but also fine filter a wide range of information to provide users with a more accurate recommendation.Finally,the user can recommend relevant scenic spots or re recommend popular scenic spots based on the filtered information.In this paper,the language model is used to transform label words into vectors,and then the vector is trained.Cosine similarity is used to calculate tags with high similarity,and emotion analysis is integrated to complete personalized recommendation of similar scenic spots.