Driving Context into Text-to-Text Privatization

$\mathit{\text{Metric}}\mathit{\text{Differential}}\mathit{\text{Privacy}}$ enables text-to-text privatization by adding calibrated noise to the vector of a word derived from an embedding space and projecting this noisy vector back to a discrete vocabulary using a nearest neighbor search. Since words are substituted without context, this mechanism is expected to fall short at finding substitutes for words with ambiguous meanings, such as $\mathit{\text{’bank’}}$. To account for these ambiguous words, we leverage a sense embedding and incorporate a sense disambiguation step prior to noise injection. We encompass our modification to the privatization mechanism with an estimation of privacy and utility. For word sense disambiguation on the $\mathit{\text{Words}}\mathit{\text{in}}\mathit{\text{Context}}$ dataset, we demonstrate a substantial increase in classification accuracy by $6.05\mathrm{\%}$.