An AI-driven framework for continuous tourist sentiment scoring using longitudinal and group-level insights with pre-trained language models (RoBERTa-CSS)

Abstract

Purpose – Tourist sentiment is typically measured as discrete categories (e.g. positive, neutral and negative) through lexicon-based or machine-learning-based approaches in extant studies. However, neuroscience and physiology scholars have argued that sentiments are continuous in nature. Treating sentiment as a categorical state may result in an overly simplified understanding of tourists’ sentiments, ultimately hindering the tourism industry’s ability to derive precise and actionable insights. This study aims to construct an AI-driven framework for continuous tourist sentiment scoring. Design/methodology/approach – This paper proposed a tool named RoBERTa-CSS (RoBERTa-based Continuous Sentiment Scoring) to calculate tourists’ continuous sentiment scores based on the pre-trained language model RoBERTa. The structure of RoBERTa is refined by adding a fully connected neural network layer, enabling the prediction of continuous sentiment scores. Using Chinese online reviews of a hotel group from multiple travel platforms, 3, 500 sentences segmented from 1, 000 randomly selected reviews were manually annotated to evaluate the proposed approach. Findings – The comparison with the state-of-the-art open-source packages, deep learning models, pre-trained language models and generative artificial intelligence tools on multiple evaluation metrics demonstrated the superiority of the proposed RoBERTa-CSS. The method was also validated on an English dataset, showing good performance. Several empirical analyses, including individual-level sentiment flow analysis, group-level sentiment distribution and longitudinal analysis, were performed using the full dataset. The results further showcased the edge of RoBERTa-CSS, compared to extant polarity categorization-oriented sentiment analysis methods. Originality/value – This study expanded the analytical ability beyond simple categorization to facilitate understanding of the complexity and diversity of human sentiment based on an improved pre-trained language model. The relevance of this paper for tourism practitioners, destination management organizations and online travel platforms is discussed. ⽬的: 在现有研究中, 游客情感通常是通过基于词典或机器学习的⽅法, 以离散类别（如积极、中性、消 极）来衡量的。然⽽, 神经科学和⽣理学学者认为, 情感在本质上是连续的。将情感视为⼀种分类状态可 能会导致对游客情感的理解过于简单化, 最终阻碍旅游业者获得精确和可操作的指导。本⽂旨在构建⼀个 ⼈⼯智能驱动的游客连续情感评分框架。 设计/⽅法/途径:本⽂提出了⼀种⼯具–RoBERTa-CSS（基于 RoBERTa 的连续情感评分）, 它是以预先训 练的语⾔模型 RoBERTa 为基础计算游客的连续情感评分。通过添加全连接神经⽹络层, RoBERTa 的结 构得到了完善, 从⽽可以预测连续情感评分。使⽤来⾃多个旅游平台的某酒店集团的中⽂消费者在线评 论, 从随机选取的 1,000 条评论中分割出 3,500 个句⼦进⾏⼈⼯标注, 以评估所提出的⽅法。 研究结果:经过与最先进的开源软件包、深度学习⼯具、预训练语⾔模型和⽣成式⼈⼯智能在多个评价指 标上的⽐较, 证明了所提出的 RoBERTa-CSS 的优越性。该⽅法还在英⽂数据集上进⾏了验证, 显⽰出良 好的性能。此外, 还利⽤整个数据集进⾏了多项实证分析, 包括个体层⾯的情感流分析、群体层⾯的情感 分布和纵向分析, 结果进⼀步显⽰了 RoBERTa-CSS 与现有的以极性分类为导向的情感分析⽅法相⽐所具 有的优势。 独创性: 本研究基于改进的预训练语⾔模型, 将分析能⼒从简单的分类扩展到了对⼈类情感复杂性和多样 性的理解。本⽂对旅游业从业⼈员、⽬的地管理机构和在线旅游平台的意义进⾏了讨论。