Scholarlink

<p>The present study addresses the methodological issue of interpreting topic models in LDA topic modeling by applying the proposed contextualization approach to the ecotourism data. This data comprises newspaper discourses on ecotourism from January 2010 to December 2023. The study underpins Computational Grounded Theory as the theoretical framework. LDA topic modeling has been applied to discover the latent topics from the corpus. The themes relevant to these latent topics are contextualized employing the endogenous and exogenous information regarding ecotourism. The contextualization approach helps interpret the models while keeping in mind the contextual dynamics. A total of 20 themes emerged from these topics. These themes were further classified under three main umbrella topics: prospects, challenges, and preservation. The prospects inform about the potential and the challenges inform about constraints regarding ecotourism, while preservation-based themes stress the need to protect tourism’s natural and cultural prospects. The study is significant as it proposes a contextualization approach to interpret the topic model. Furthermore, by applying this approach, the study offers insight into the dynamics of ecotourism in Pakistan.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Addressing the interpretability of big data for social sciences: a contextualization approach to LDA topic models

  • 期刊论文

  • 发表日期：

    2025年12月11日

查看全文
  • Fasih Ahmed

摘要

The present study addresses the methodological issue of interpreting topic models in LDA topic modeling by applying the proposed contextualization approach to the ecotourism data. This data comprises newspaper discourses on ecotourism from January 2010 to December 2023. The study underpins Computational Grounded Theory as the theoretical framework. LDA topic modeling has been applied to discover the latent topics from the corpus. The themes relevant to these latent topics are contextualized employing the endogenous and exogenous information regarding ecotourism. The contextualization approach helps interpret the models while keeping in mind the contextual dynamics. A total of 20 themes emerged from these topics. These themes were further classified under three main umbrella topics: prospects, challenges, and preservation. The prospects inform about the potential and the challenges inform about constraints regarding ecotourism, while preservation-based themes stress the need to protect tourism’s natural and cultural prospects. The study is significant as it proposes a contextualization approach to interpret the topic model. Furthermore, by applying this approach, the study offers insight into the dynamics of ecotourism in Pakistan.