An important aspect of human resource management is understanding employees' values, motivations, and factors of satisfaction. Relationships between employee satisfaction, values, positivity and negativity, employee retention, and industry allow human resource managers to make decisions on employee recruitment techniques and budgets, employee retention efforts, and even company image initiatives. Traditional methods for identifying employee values, job satisfaction, and linkages to important company outcomes are through employee entrance/exit surveys, annual evaluations, and informal interactions. In recent years, companies and researchers have begun to take advantage of the increasing ubiquity of online reviews, including unstructured text responses. In this paper we explore 682,176 employee reviews of Fortune 50 companies from Indeed.com using Structural Topic Modeling (STM) to identify salient aspects in employee reviews and automatically infer latent topics that are salient factors related to employee satisfaction. We compare these topics with the five facets of the Job Descriptive Index (JDI) that are traditionally used to measure job satisfaction among human resource experts. We conduct further analysis comparing comments of current vs. former employees. Results suggests that management may be an important salient factor in employee satisfaction and turnover, as it is consistently mentioned by current and former employees in both positive and the negative comments. Similarly, monetary benefits are mentioned as a negative by former employees, but as a positive for current employees. Furthermore, we examine topic differences by industry sector and find that retail employees frequently mention Pay and Benefits and Length of Breaks, whereas technology sector employees were more concerned with Work-Life Balance. Our results can be used to support company behavioral management decision makers to conceive and evaluate initiatives intended to enhance employee satisfaction. Furthermore, our techniques, including the visualization of topic composition and quality, are generalizable to any setting that uses topic discovery from unstructured text, such as customer reviews, social media posts, and critic reviews, especially when comparing topics across entities.

人力资源管理的一个重要方面是了解员工的价值观、动机和满意度因素。员工满意度、价值观、积极性和消极性、员工保留率和行业之间的关系使人力资源经理能够就员工招聘技术和预算、员工保留工作甚至公司形象举措做出决策。识别员工价值观、工作满意度以及与重要公司成果的联系的传统方法是通过员工入职/离职调查、年度评估和非正式互动。近年来，公司和研究人员开始利用日益普遍的在线评论，包括非结构化文本回复。在本文中，我们探讨了来自 Indeed 的财富 50 强公司的 682,176 条员工评论。com 使用结构主题建模 (STM) 来识别员工评论中的显着方面，并自动推断与员工满意度相关的显着因素的潜在主题。我们将这些主题与传统上用于衡量人力资源专家工作满意度的工作描述指数 (JDI) 的五个方面进行了比较。我们对现任和前任员工的评论进行了进一步分析。结果表明 我们对现任和前任员工的评论进行了进一步分析。结果表明 我们对现任和前任员工的评论进行了进一步分析。结果表明管理可能是影响员工满意度和离职率的一个重要的突出因素，因为现任和前任员工在正面和负面评论中都不断提到这一点。同样，货币福利被前员工认为是负面的，但对现任员工来说是积极的。此外，我们检查了行业部门的主题差异，发现零售员工经常提到薪酬福利和休息时间，而技术部门员工更关心工作与生活的平衡. 我们的结果可用于支持公司行为管理决策者构思和评估旨在提高员工满意度的举措。此外，我们的技术，包括主题组成和质量的可视化，可推广到任何使用非结构化文本主题发现的设置，例如客户评论、社交媒体帖子和评论家评论，尤其是在跨实体比较主题时。