Business Analytics in Tourism: Uncovering Knowledge from Crowds

Abstract Business Analytics leverages value from data, thus being an important tool for the decision-making process. However, the presence of data in different formats is a new challenge for analysis. Textual data has been drawing organizational attention as thousands of people express themselves daily in text, like the description of customer perceptions in the tourism and hospitality area. Despite the relevance of customer data in textual format to support decision making of hotel managers, its use is still modest, given the difficulty of analyzing and interpreting the large amounts of data. Our objective is to identify the main evaluation topics presented in online guest reviews and reveal changes throughout the years. We worked with 23,229 hotel reviews collected from TripAdvisor website through WebScrapping packages in R, and used a text mining approach (Latent Semantic Analysis) to analyze the data. This contributes with practical implications to hotel managers by demonstrating the applicability of text data and tools based on open-source solutions and by providing insights about the data and assisting in the decision-making process. This article also contributes in presenting a stepwise text analysis, including capturing, cleaning and formatting publicly available data for organizational specialists.

摘要：商业分析通过挖掘数据价值，成为决策流程中的重要工具。然而，多格式数据的存在为分析工作带来了全新挑战。随着每日海量用户以文本形式表达观点——例如旅游与酒店行业中顾客感知的描述文本——文本数据正愈发受到组织机构的关注。尽管顾客文本数据对酒店管理者的决策支持具备重要价值，但由于难以分析和解读海量此类数据，其实际应用仍较为有限。本研究旨在识别在线宾客评论中的核心评价主题，并揭示其随时间推移的变化趋势。我们借助R语言中的网络爬虫（Web Scraping）工具包，从猫途鹰（TripAdvisor）网站采集了23229条酒店评论，并采用文本挖掘方法结合潜在语义分析（Latent Semantic Analysis）对数据进行分析。本研究通过展示基于开源方案的文本数据与工具的适用性，同时提供数据洞察以辅助决策流程，为酒店管理者提供了实践参考价值。此外，本文还提出了一套分步式文本分析流程，涵盖公开可用数据的采集、清洗与格式化，以供组织内专业人员使用。