Text mining and dimension reduction method application into exploring isomorphic pressures in corporate communication on textual tweet data about sustainability in the energy sector

Abstract

The study analyses the isomorphism pressures within the context of sustainability by exploring the Twitter communication in the energy sector. Recently, there can be observed the increasing focus on interactive and communicative construction of an institution to understand how the organizations sustain the institutional pressures. The rhetorical commitments that create narrative dynamics in organizational communication are central to institutional diffusion and change. Social Media, Twitter, in particular, has been demonstrated as the new opportunity to explore the linguistic dimension in corporate communications. We propose the use of Social Media linguistic data (tweets with their hashtags and keywords) and the triangulated method (text mining, web mining, and linguistic and content analysis) to examine the tweets´ trends in each company. Based on the institutional theory of organizational communication, the paper examines the relation between the idea of sustainability and isomorphism that leads to the adoption of similar models and attitudes among the organizations. It applies the text mining and correspondence methods within the R software. The energy sector tweets in English (from 2016) were treated by the text mining processes of the statistical linguistic analysis in the R tool. Text mining, involving the linguistic, statistical, and the machine learning techniques reveals and visualizes the latent structures of the content in an unstructured or weakly structured text data in a given collection of documents.