Toward Micro-Dialect Identification in Diaglossic and Code-Switched Environments

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, pages 5855–5876, November 16–20, 2020. c⃝2020 Association for Computational Linguistics 5855 Toward Micro-Dialect Identification in Diaglossic and Code-Switched Environments Muhammad Abdul-Mageed Chiyu Zhang AbdelRahim Elmadany Lyle Ungar† Natural Language Processing Lab, University of British Columbia †Computer and Information Science, University of Pennsylvania {muhammad.mageed,a.elmadany}@ubc.ca chiyuzh@mail.ubc.ca †ungar@cis.upenn.edu Abstract Although prediction of dialects is an important language processing task, with a wide range of applications, existing work is largely limited to coarse-grained varieties. Inspired by geolo- cation research, we propose the novel task of Micro-Dialect Identification (MDI) and intro- duce MARBERT, a new language model with striking abilities to predict a fine-grained vari- ety (as small as that of a city) given a single, short message. For modeling, we offer a range of novel spatially and linguistically-motivated multi-task learning models. To showcase the utility of our models, we introduce a new, large-scale dataset of Arabic micro-varieties (low-resource) suited to our tasks. MARBERT predicts micro-dialects with 9.9% F1, ∼ 76× better than a majority class baseline. Our new language model also establishes new state-of- the-art on several external tasks.1 1 Introduction Sociolinguistic research has shown how language varies across geographical regions, even for areas as small as different parts of the same city (Labov, 1964; Trudgill, 1974). These pioneering studies often used field work data from a handful of indi- viduals and focused on small sets of carefully cho- sen features, often phonological. Inspired by this early work, researchers have used geographically tagged social media data from hundreds of thou- sands of users to predict user location (Paul and Dredze, 2011; Amitay et al., 2004; Han et al., 2014; Rahimi et al., 2017; Huang and Carley, 2019b; Tian et al., 2020; Zhong et al., 2020) or to develop lan- guage identification tools (Lui and Baldwin, 2012; Zubiaga et al., 2016; Jurgens et al., 2017a; Dunn and Adams, 2020). Whether it is possible at all to predict the micro-varieties 2 of the same general 1Our labeled data and models will be available at: https: //github.com/UBC-NLP/microdialects. 2We use micro-variety and micro-dialect interchangeably. language is a question that remains, to the best of our knowledge, unanswered. In this work, our goal is to investigate this specific question by introduc- ing the novel task of Micro-Dialect Identification (MDI). Given a single sequence of characters (e.g., a single tweet), the goal of MDI is to predict the particular micro-variety (defined at the level of a city) of the community of users to which the post- ing user belongs. This makes MDI different from geolocation in at least two ways: in geolocation, (1) a model consumes a collection of messages (e.g., 8-85 messages in popular datasets (Huang and Carley, 2019a) and (2) predicts the location of the posting user (i.e., user-level). In MDI, a model takes as input a single message, and predicts the micro-variety of that message (i.e., message-level). While user location and micro-dialect (MD) are conceptually related (e.g., with a tag such as Seat- tle for …

Content truncated to 500 words