Uncovering Limitations of Large Language Models in Information Seeking from Tables
2024 Findings of ACL [Download]
We propose a benchmark dataset demonstrating that LLMs are not good at seeking information from tables.
2024 Findings of ACL [Download]
We propose a benchmark dataset demonstrating that LLMs are not good at seeking information from tables.
2023 EMNLP [Download]
We propose to learn the task description (guideline) to learn from training samples for in-context learning
2023 NeurIPS [Download]
We prove that, under some assumptions, ensemble model outperforms its members on a range of coverage for selective classification.
2022 KDD [Download]
Extracting the meaning of numbers in complex structured tables.
2021 JCST [Download]
We study the problem of extracting variable-depth “logical document hierarchy”, i.e. table of contents, from long documents.
2021 CIKM [Download]
Assess compliance for annual reports of listed company against hundreds of listing rules, and find document structure information is important.
2021 KDD [Download]
Extracting numerical calculation formulas in tables.
2020 AAAI [Download]
Solve math word problem as extracting a DAG structure upon text.
2020 KDD [Download]
Automatic cross-checking over numerical facts in tables in financial documents.
2019 CIKM [Download]
Nested relation extraction with iterative neural network to extract structured information.
2018 WWW [Download]
Extract formulas from text using deep learning method.