Saved in:
Bibliographic Details
Main Authors: Li, Yinghao, Miao, Siyu, Huang, Heyan, Gao, Yang
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2406.14828
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929393651154944
author Li, Yinghao
Miao, Siyu
Huang, Heyan
Gao, Yang
author_facet Li, Yinghao
Miao, Siyu
Huang, Heyan
Gao, Yang
contents Domain adaptation aims to enable Large Language Models (LLMs) to generalize domain datasets unseen effectively during the training phase. However, factors such as the size of the model parameters and the scale of training data are general influencers and do not reflect the nuances of domain adaptation performance. This paper investigates the fine-grained factors affecting domain adaptation performance, analyzing the specific impact of `words' in training data on summarization tasks. We propose quantifying dataset learning difficulty as the learning difficulty of generative summarization, which is determined by two indicators: word-based compression rate and abstraction level. Our experiments conclude that, when considering dataset learning difficulty, the cross-domain overlap and the performance gain in summarization tasks exhibit an approximate linear relationship, which is not directly related to the number of words. Based on this finding, predicting a model's performance on unknown domain datasets is possible without undergoing training.
format Preprint
id arxiv_https___arxiv_org_abs_2406_14828
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Word Matters: What Influences Domain Adaptation in Summarization?
Li, Yinghao
Miao, Siyu
Huang, Heyan
Gao, Yang
Computation and Language
Domain adaptation aims to enable Large Language Models (LLMs) to generalize domain datasets unseen effectively during the training phase. However, factors such as the size of the model parameters and the scale of training data are general influencers and do not reflect the nuances of domain adaptation performance. This paper investigates the fine-grained factors affecting domain adaptation performance, analyzing the specific impact of `words' in training data on summarization tasks. We propose quantifying dataset learning difficulty as the learning difficulty of generative summarization, which is determined by two indicators: word-based compression rate and abstraction level. Our experiments conclude that, when considering dataset learning difficulty, the cross-domain overlap and the performance gain in summarization tasks exhibit an approximate linear relationship, which is not directly related to the number of words. Based on this finding, predicting a model's performance on unknown domain datasets is possible without undergoing training.
title Word Matters: What Influences Domain Adaptation in Summarization?
topic Computation and Language
url https://arxiv.org/abs/2406.14828