Mapping the World of Consumption: Computational Linguistics Analysis of the Google Text Corpus

Abstract

This article describes a method that develops overviews to bring out the relationships between any loosely connected set of actors/objects. The study examines 37 principal actors involved in the processes of consumption (consumers, brands, ads, stores…), and how they are described on the internet in the Google corpus of linguistic data. The verbs used with each actor constitute a profile of the behaviors that people ascribe to that actor. The analysis synthesizes these profiles into pictures using multidimensional scaling. Separate analyses examine actors as (a) the subject of the verbs, and (b) the object of the verbs. This reliability check reveals highly congruent pictures of the relationship between actors. The paper subsequently examines the most distinctive behaviors of contrasting actors to further understand selected parts of the picture (e.g., how products differ from services). Web chatter is unrestricted in topic, which is produced by people and for people. Therefore, the corpus is a rich source of data, not just for marketing research - as illustrated here - but for almost any branch of research into human affairs.

Keywords: consumption, mapping, Google, verbs