Understanding "Democratization" in NLP and ML Research
This addresses conceptual clarity for researchers in NLP/ML regarding a frequently used but inconsistently defined term, though it is incremental as it analyzes existing discourse rather than proposing new methods.
The paper analyzed how the term 'democratization' is used in NLP and ML research through large-scale mixed-methods analyses of publications, finding it typically refers to technology access without engaging with democratic theory, while other uses are more theory-grounded.
Recent improvements in natural language processing (NLP) and machine learning (ML) and increased mainstream adoption have led to researchers frequently discussing the "democratization" of artificial intelligence. In this paper, we seek to clarify how democratization is understood in NLP and ML publications, through large-scale mixed-methods analyses of papers using the keyword "democra*" published in NLP and adjacent venues. We find that democratization is most frequently used to convey (ease of) access to or use of technologies, without meaningfully engaging with theories of democratization, while research using other invocations of "democra*" tends to be grounded in theories of deliberation and debate. Based on our findings, we call for researchers to enrich their use of the term democratization with appropriate theory, towards democratic technologies beyond superficial access.