Blog post #3 - Research Gap
Research GAP
While research has examined the benefits of using AI in business writing, the academic field, and language-to-language translation, research is lacking in the area of AI translation specifically for converting "legalese" to plain language. Although Moneus & Sahari (2024) have conducted research on translating legal documents into other languages, gaps remain in the current body of literature regarding the translation of "legalese" into plain language.
The questions I would like to answer are: Can AI effectively transform legal language to plain language addressing three main factors of confusion in legal documents, namely passive voice, low-frequency words, and confusing sentence structures? Can it achieve this to the satisfaction of both the legal writer and the reader?
In my other class, I had to lead a discussion (yesterday) for a chapter on how writing has to be simplified by technical writers before it's translated into another language (by a human or a machine/software). While this chapter was more focused on translating technical writing in the form of manuals/instructions into plain language to "prep it" for translation, I wonder if research has been done on that? (My gut says it's less politicized and consumer-oriented so work has probably been done.) One of the big concepts was GILT (Globalization Internationalization Localization Translation) for products and services. Maybe it could be triangulated into your project? One of the big pitfalls for language translation was that the proofreading involved was very surface level (they looked for typos rather than confirming whether the instructions were correct). I don't know if this same kind of issue would happen for AI translating legal documents, but it's something to consider.
ReplyDeleteI enjoy reading about your project, it always gets me thinking. Looking forward to your discoveries! Let me know if you're interested in that chapter.
ReplyDeleteI guess someone would need to create an algorithm to teach the AI how to make the translation. I learned about supervised learning when creating algorithms for my Literature review. I think someone must program the AI using a supervised learning algorithm to generate those translation capabilities. Are you planning to use a qualitative method to answer your questions? Will you create a survey?