Person ED7526
- Interviewed on 2023-06-23
- Currently works at Grammarly
Computational linguists use NLP technology to make Grammarly suggestions. They focus on production readiness and use a dictionary, Ngram, pipeline, ML, and DL.
Currently, we are focusing on measuring the quality of LLM's results through the development of metrics and error correction methods. For example, let's say we need to transform impolite text into polite text through research on politeness. There will be many subtasks and approaches, such as defining subtasks and considering the differences in politeness between British and American English.
When considering the three roles, there are distinct differences. A Researcher may have limited knowledge of Linguistics and typically works with deep learning. An Analytical Linguist is involved in annotation projects but does not focus on writing production code. On the other hand, a Computational Linguist is capable of programming for various NLP tasks, handling metrics and Linguistics all at once.
ACL, the association of computational linguistics, launched an ACL 60-60 initiative. The initiative aims to facilitate scientific communication in various modalities, including text, speech, and sign language, and across 60 different languages.
Personally have an interest in semantics, especially in AMR.