Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Has It All Been Solved? Open NLP Research Questions Not Solved by Large Language Models

Oana Ignat, Zhijing Jin, Artem Abzaliev, Laura Biester, Santiago Castro, Naihao Deng, Xinyi Gao, Aylin Gunal, Jacky He, Ashkan Kazemi, Muhammad Khalifa, Namho Koh, Andrew Lee, Siyang Liu, Do June Min, Shinka Mori, Joan Nwatu, Veronica Perez-Rosas, Siqi Shen, Zekun Wang, Winston Wu, Rada Mihalcea

TL;DR

This paper identifies fourteen different research areas encompassing 45 research directions that require new research and are not directly solvable by LLMs, and identifies areas currently addressed by LLMs, but where LLMs lag behind in performance or those focused on LLM development.

Abstract

Recent progress in large language models (LLMs) has enabled the deployment of many generative NLP applications. At the same time, it has also led to a misleading public discourse that ``it's all been solved.'' Not surprisingly, this has, in turn, made many NLP researchers -- especially those at the beginning of their careers -- worry about what NLP research area they should focus on. Has it all been solved, or what remaining questions can we work on regardless of LLMs? To address this question, this paper compiles NLP research directions rich for exploration. We identify fourteen different research areas encompassing 45 research directions that require new research and are not directly solvable by LLMs. While we identify many research areas, many others exist; we do not cover areas currently addressed by LLMs, but where LLMs lag behind in performance or those focused on LLM development. We welcome suggestions for other research directions to include: https://bit.ly/nlp-era-llm

Has It All Been Solved? Open NLP Research Questions Not Solved by Large Language Models

TL;DR

This paper identifies fourteen different research areas encompassing 45 research directions that require new research and are not directly solvable by LLMs, and identifies areas currently addressed by LLMs, but where LLMs lag behind in performance or those focused on LLM development.

Abstract

Recent progress in large language models (LLMs) has enabled the deployment of many generative NLP applications. At the same time, it has also led to a misleading public discourse that ``it's all been solved.'' Not surprisingly, this has, in turn, made many NLP researchers -- especially those at the beginning of their careers -- worry about what NLP research area they should focus on. Has it all been solved, or what remaining questions can we work on regardless of LLMs? To address this question, this paper compiles NLP research directions rich for exploration. We identify fourteen different research areas encompassing 45 research directions that require new research and are not directly solvable by LLMs. While we identify many research areas, many others exist; we do not cover areas currently addressed by LLMs, but where LLMs lag behind in performance or those focused on LLM development. We welcome suggestions for other research directions to include: https://bit.ly/nlp-era-llm
Paper Structure (68 sections, 2 figures, 1 table)

This paper contains 68 sections, 2 figures, 1 table.

Table of Contents

  1. Background
  2. About This Document.
  3. Disclaimer.
  4. Document Organization.
  5. Fundamental NLP
  6. Multilinguality
  7. Low-resource machine translation.
  8. Multilingual models that work well for all languages.
  9. Code-switching.
  10. Reasoning
  11. Complex reasoning.
  12. Responsible reasoning in social contexts.
  13. Formally defining reasoning and designing a proper evaluation framework.
  14. Knowledge Bases
  15. Automatic knowledge base construction.
  16. ...and 53 more sections

Figures (2)

  • Figure 1: Research areas in NLP that are rich for exploration, spanning Fundamental, Responsible, and Applied research.
  • Figure 2: So "What should I work on?" Based on your motivation, you can choose from at least three main NLP areas: Fundamental, Responsible, and Applied. Furthermore, your choice could depend on whether or not you like fast-paced research, whether you want to make models transparent and trustworthy, or, whether you prefer broader or more specific domains.