Lörres, Möppes, and the Swiss. (Re)Discovering regional patterns in anonymous social media data

Abstract

We study regional similarities and differences in language use on an anonymous mobile chat application in the German-speaking area. We use a neural network on 2.3 million online conversations to automatically learn representations of words and cities. These linguistic-use-based representations capture regional distinctions in a high-dimensional vector space that can be clustered and visualized to discover patterns in the data. We find that the resulting regional patterns are closely linked to the traditional division of German dialects, even though most of the conversations are written in standard German. The resulting maps correspond to traditional dialect divisions and language-external spatial structures, with a few notable exceptions that can be explained through external factors.Our method also facilitates two qualitative analyses, allowing us to discover geographically-pertinent words for various regional levels, as well as creating regional group-specific style profiles based on various linguistic resources. The results of our study strongly suggest the existence of region-specific patterns of language use (“digital regiolects”) representing distinctive strategies of linguistic stylization in relation to linguistic resources and topics. As a methodological contribution, we show how linguistic theory can drive the application and direction of neural network-based representation learning, and how their judicious application provides the basis for qualitative analysis of large-scale data collections.

Raumgeographische Verteilung von Twitter-Hashtags im deutschen Sprachraum

“…Auf diese Weise werden bestimmte Hashtags (solche, die lokal geclustert sind, aber keine eindeutigen geografischen oder anderen toponomastischen Merkmale bezeichnen) mit Orten assoziiert. Diese Indexikalität und Markierung wird für lokale Gemein-schaften konstitutiv, wenn sie in globalen Netzwerken wie dem Internet vertreten sind, und kann in gewisser Weise der Art und Weise entsprechen, in der bestimmte lexikalische Elemente mit bestimmten lokalen Dialekten in der traditionellen Dialektologie assoziiert werden (Purschke & Hovy 2019).…”

Large Language Models as Computational Linguistics Tools: A Comparative Analysis of ChatGPT and Google Machine Translations

“…Computational sociolinguists employ computational techniques to gauge the interplay between society and language, encompassing the stylistic and structural attributes that differentiate speakers [16]. Language variation is intricately tied to social identity [17], spanning group affiliation [18], geographic location [19], and socioeconomic status [20][21], as well as individual characteristics such as age and gender [22].…”

Attitudes Toward Multilingualism in Luxembourg. A Comparative Analysis of Online News Comments and Crowdsourced Questionnaire Data

“…As a consequence, the applicability and appropriateness of many NLP applications for everyday language is still limited, despite the great advances in computer science and AI research (Bender and Koller, 2020 ). In recent years, there has been a new trend in the NLP community that is increasingly concerned with language as a social phenomenon and that tries to incorporate sociolinguistic knowledge into the analysis of data and the development of new tools and models (Broadwell et al, 2013 ; Eisenstein, 2015 ; Nguyen, 2017 ; Purschke and Hovy, 2019 ).…”