•  
  •  
 
Communications of the Association for Information Systems

Author ORCID Identifier

Asish Satpathy: 0000-0001-7426-2909

Michael A. Erskine: 0000-0002-8802-7624

Abstract

Text-based analytics, including sentiment analysis and topic modeling, are widely used in Information Systems (IS) research to analyze user-generated content. Although these methods effectively summarize evaluative language and thematic patterns at scale, they often treat textual observations as aspatial, even when digital interactions are explicitly linked to physical locations. This limitation constrains interpretation when participation, expectations, and evaluations vary systematically across neighborhoods and regions.

This tutorial presents a five-stage geo-enriched text analytics design process that integrates spatial context into established IS text analytics workflows. It clarifies and formalizes key design decisions concerning the unit of analysis, spatial scale, contextual data alignment, feature integration, and model selection, without introducing new algorithms. The approach is illustrated using Yelp Open Dataset reviews of Starbucks locations in the Greater Philadelphia area, demonstrating how location-linked text can be enriched with neighborhood-level socioeconomic and demographic attributes through a transparent and replicable workflow. Sensitivity analyses across alternative buffer sizes highlight spatial scale as a substantive modeling decision, while spatial diagnostics illustrate when global models provide stable, interpretable summaries.

The primary value is methodological and pedagogical. This tutorial provides a reusable template for embedding geographic context into text analytics, enabling transparent, context-sensitive interpretation in IS research and supporting structured skill development within IS community.

Download

Share

COinS
 

When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.