在体验经济时代下,人们对旅游服务体验的诉求不断提升。由于线上公众意见能够更加真实地表达游客的体验和反馈,越来越多的消费者依赖公众分享的信息辅助进行旅行决策。在旅行决策中,消费者的感知价值发挥着关键作用。传统基于用户内容的推荐方法研究多从用户行为偏好视角进行研究,忽视了用户感知价值的作用,影响了旅游服务的个性化推荐效果。因此本文从感知价值视角出发,提出适用于公众意见的用户感知价值评估方法。首先,将用户期望作为前景理论参照点,通过词性抽取规则和半监督学习方法,有效解决了公众文本中用户期望信息稀缺的问题;其次,提出融合用户期望的群体聚类优化方法,提升了群体期望构建的准确性。进而,将前景理论和多属性决策模型结合评估用户感知价值。通过概率语言决策矩阵刻画公众意见,基于前景理论构建概率语言感知矩阵,将公众意见转化为感知价值。以TODIM方法为基础,集结大群体公众意见得到备选方案的量化评估。最后,基于真实旅游评论数据的实证研究验证了该方法的有效性,为提升个性化推荐效果提供了新思路。In the era of experience economy, consumers’ demand for travel service experience is rising. Online public opinion can express tourists’ experiences and feedback more objectively, and more and more consumers rely on the information shared by the public to assist in their travel decisions. Consumers’ perceived value plays a key role in travel decision-making. The traditional user content-based recommendation methods are mostly studied from the perspective of user behavioral preferences, ignoring the influence of user perceived value, which affects the effect of personalized recommendation of travel services. Therefore, this paper proposes a user perceived value assessment method applicable to public opinion from the perspective of perceived value. Firstly, th
