Saved in:
Bibliographic Details
Main Authors: Mai, Quan, Gauch, Susan, Adams, Douglas
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2406.17282
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916301250756608
author Mai, Quan
Gauch, Susan
Adams, Douglas
author_facet Mai, Quan
Gauch, Susan
Adams, Douglas
contents We introduce SetBERT, a fine-tuned BERT-based model designed to enhance query embeddings for set operations and Boolean logic queries, such as Intersection (AND), Difference (NOT), and Union (OR). SetBERT significantly improves retrieval performance for logic-structured queries, an area where both traditional and neural retrieval methods typically underperform. We propose an innovative use of inversed-contrastive loss, focusing on identifying the negative sentence, and fine-tuning BERT with a dataset generated via prompt GPT. Furthermore, we demonstrate that, unlike other BERT-based models, fine-tuning with triplet loss actually degrades performance for this specific task. Our experiments reveal that SetBERT-base not only significantly outperforms BERT-base (up to a 63% improvement in Recall) but also achieves performance comparable to the much larger BERT-large model, despite being only one-third the size.
format Preprint
id arxiv_https___arxiv_org_abs_2406_17282
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle SetBERT: Enhancing Retrieval Performance for Boolean Logic and Set Operation Queries
Mai, Quan
Gauch, Susan
Adams, Douglas
Computation and Language
We introduce SetBERT, a fine-tuned BERT-based model designed to enhance query embeddings for set operations and Boolean logic queries, such as Intersection (AND), Difference (NOT), and Union (OR). SetBERT significantly improves retrieval performance for logic-structured queries, an area where both traditional and neural retrieval methods typically underperform. We propose an innovative use of inversed-contrastive loss, focusing on identifying the negative sentence, and fine-tuning BERT with a dataset generated via prompt GPT. Furthermore, we demonstrate that, unlike other BERT-based models, fine-tuning with triplet loss actually degrades performance for this specific task. Our experiments reveal that SetBERT-base not only significantly outperforms BERT-base (up to a 63% improvement in Recall) but also achieves performance comparable to the much larger BERT-large model, despite being only one-third the size.
title SetBERT: Enhancing Retrieval Performance for Boolean Logic and Set Operation Queries
topic Computation and Language
url https://arxiv.org/abs/2406.17282