Saved in:
Bibliographic Details
Main Authors: Kumar, Rohan, Kim, Youngmin, Ravi, Sunitha, Sun, Haitian, Faloutsos, Christos, Salakhutdinov, Ruslan, Yoon, Minji
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2403.01382
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917602915254272
author Kumar, Rohan
Kim, Youngmin
Ravi, Sunitha
Sun, Haitian
Faloutsos, Christos
Salakhutdinov, Ruslan
Yoon, Minji
author_facet Kumar, Rohan
Kim, Youngmin
Ravi, Sunitha
Sun, Haitian
Faloutsos, Christos
Salakhutdinov, Ruslan
Yoon, Minji
contents Pretrained Large Language Models (LLMs) have gained significant attention for addressing open-domain Question Answering (QA). While they exhibit high accuracy in answering questions related to common knowledge, LLMs encounter difficulties in learning about uncommon long-tail knowledge (tail entities). Since manually constructing QA datasets demands substantial human resources, the types of existing QA datasets are limited, leaving us with a scarcity of datasets to study the performance of LLMs on tail entities. In this paper, we propose an automatic approach to generate specialized QA datasets for tail entities and present the associated research challenges. We conduct extensive experiments by employing pretrained LLMs on our newly generated long-tail QA datasets, comparing their performance with and without external resources including Wikipedia and Wikidata knowledge graphs.
format Preprint
id arxiv_https___arxiv_org_abs_2403_01382
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Automatic Question-Answer Generation for Long-Tail Knowledge
Kumar, Rohan
Kim, Youngmin
Ravi, Sunitha
Sun, Haitian
Faloutsos, Christos
Salakhutdinov, Ruslan
Yoon, Minji
Computation and Language
Pretrained Large Language Models (LLMs) have gained significant attention for addressing open-domain Question Answering (QA). While they exhibit high accuracy in answering questions related to common knowledge, LLMs encounter difficulties in learning about uncommon long-tail knowledge (tail entities). Since manually constructing QA datasets demands substantial human resources, the types of existing QA datasets are limited, leaving us with a scarcity of datasets to study the performance of LLMs on tail entities. In this paper, we propose an automatic approach to generate specialized QA datasets for tail entities and present the associated research challenges. We conduct extensive experiments by employing pretrained LLMs on our newly generated long-tail QA datasets, comparing their performance with and without external resources including Wikipedia and Wikidata knowledge graphs.
title Automatic Question-Answer Generation for Long-Tail Knowledge
topic Computation and Language
url https://arxiv.org/abs/2403.01382