Research on Hotel Review Text Mining and Application under Machine Learning and LDA Model
Lu Zhang, Zhaoyin Ding
Anhui Sanlian University
阅读操作
确认中在文库中上传 PDF 后可生成中文音频讲解。
摘要与影响
Online hotel reviews capture what guests truly value, but their sheer volume and unstructured language make it difficult for managers to identify the most influential satisfaction drivers in a timely, systematic way. This paper aims to uncover the key factors shaping hotel user satisfaction by applying Latent Dirichlet Allocation (LDA) topic modeling to a large corpus of real-world hotel review texts. We build a complete text-mining pipeline that includes data cleaning, Chinese word segmentation, stop-word removal, and bag-of-words representation, followed by LDA training and topic-number selection using quantitative quality indicators and interpretability checks. To make the results transparent and manager-friendly, we further visualize topic structure and term relevance through LDA-specific visual analytics, enabling intuitive inspection of both overall term frequency and topic-specific salience. The extracted topics consistently organize around concrete service and product dimensions that matter to guests, such as room cleanliness and hygiene, staff attitude and service responsiveness, location and transportation convenience, facility condition and maintenance, noise and sleep quality, and perceived value for money. By translating these latent themes into actionable operational levers, the study offers an evidence-based perspective on where hotels should prioritize improvements to raise satisfaction and reduce complaints. The proposed framework is lightweight, reproducible, and readily transferable to other hospitality scenarios where decision-makers need interpretable insights from large-scale review text.
逐年被引趋势
暂无年度引用数据
关键指标
同类平均 = 1
同领域 · 同年份 · 同类型
Google Scholar 与 OpenAlex 的被引统计范围不同,数值存在差异属正常。
AI 辅助阅读
依据:摘要
可就本文提问;依据不足时会说明。
学术脉络
学科主题
计算机 / AISentiment Analysis and Opinion Mining
Computational and Text Analysis Methods · Topic Modeling