Danielle SaundersBlogging the thesis entry 4: Incredibly simple domain adaptation (and its problems)In a series of posts summarizing my PhD thesis, “Domain adaptation for neural machine translation”, here I talk through work on simply…Aug 2, 2021Aug 2, 2021
Danielle SaundersBlogging the thesis entry 3: How are people adapting machine translation?In a series of posts summarizing my PhD thesis, “Domain adaptation for neural machine translation”, here I review existing ways to adapt…Mar 3, 2021Mar 3, 2021
Danielle SaundersBlogging the thesis entry 2: Neural machine translation from start to finishIn a series of posts summarizing my PhD thesis, “Domain adaptation for neural machine translation”, this post covers my first literature…Feb 22, 2021Feb 22, 2021
Danielle SaundersBlogging the thesis entry 1: The introductionI’ve decided to summarize my PhD thesis, “Domain adaptation for neural machine translation”, in a series of blog posts. In this post on…Feb 13, 2021Feb 13, 2021
Danielle SaundersBlogging the thesis entry 0: about the project(s)I’ve decided to summarize my PhD thesis, “Domain adaptation for neural machine translation”, in a series of blog posts. Here, I talk about…Feb 5, 2021Feb 5, 2021
Danielle SaundersMachine translation doesn’t translate gender right — unless you make itMachine translation struggles to translate people’s gender, especially for sentences referring to multiple people. In a previous post, we…Jan 29, 2021Jan 29, 2021
Danielle SaundersinAnalytics VidhyaMitigating gender bias in machine translation: adapt, then correctMachine translation struggles to translate people’s gender. Instead, it reflects biases from training data. We explore ways to unlearn…Jan 24, 2021Jan 24, 2021