Back

Bridging Domains in Chronic Lower Back Pain: Large Language Models and Ontology-driven Strategies for Knowledge Graph Construction

Anderson, P. E.; Davidson, J. M.; Lin, D.; Migler, T.; Ho, I.; Koenig, C.; Bittner, M.; Kaplan, S.; Paraiso, M.; Buhn, N.; Stokes, E.; Hunt, T.; Ropella, G.; Lotz, J.

2024-03-14 bioinformatics
10.1101/2024.03.11.584505 bioRxiv
Show abstract

Link prediction and entity resolution play pivotal roles in uncovering hidden relationships within networks and ensuring data quality in the era of heterogeneous data integration. This paper explores the utilization of large language models to enhance link prediction, particularly through knowledge graphs derived from transdisciplinary literature. Investigating zero-shot entity resolution techniques, we examine the impact of ontology-based and large language model approaches on the stability of link prediction results. Through a case study focusing on chronic lower back pain research, we analyze workflow decisions and their influence on prediction outcomes. Our research underscores the importance of robust methodologies in improving predictive accuracy and data integration across diverse domains.

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

1
PLOS ONE
● 5266 papers in training set
Top 21%
7.8%
Similar papers in this journal
2
BMC Medical Informatics and Decision Making
○ 43 papers in training set
Top 0.2%
7.2%
Similar papers in this journal
3
Journal of Biomedical Informatics
○ 47 papers in training set
Top 0.2%
6.7%
Similar papers in this journal
4
Bioinformatics Advances
◐ 203 papers in training set
Top 0.7%
6.7%
5
Bioinformatics
◐ 1204 papers in training set
Top 4%
6.7%
Similar papers in this journal
6
GigaScience
◐ 212 papers in training set
Top 0.3%
6.7%
Similar papers in this journal
7
Journal of the American Medical Informatics Association
◐ 71 papers in training set
Top 0.7%
5.1%
8
BMC Bioinformatics
○ 457 papers in training set
Top 2%
4.3%
50% of probability mass above
9
JAMIA Open
◐ 42 papers in training set
Top 0.5%
4.0%
10
PLOS Computational Biology
● 1863 papers in training set
Top 9%
4.0%
11
Database
◐ 61 papers in training set
Top 0.2%
3.2%
12
Scientific Reports
○ 3612 papers in training set
Top 43%
2.4%
13
BioData Mining
○ 22 papers in training set
Top 0.2%
2.1%
14
Computers in Biology and Medicine
○ 128 papers in training set
Top 2%
1.9%
15
Artificial Intelligence in Medicine
○ 17 papers in training set
Top 0.4%
1.7%
16
Computational and Structural Biotechnology Journal
● 242 papers in training set
Top 3%
1.7%
17
Scientific Data
○ 209 papers in training set
Top 1%
1.7%
18
Frontiers in Bioinformatics
○ 49 papers in training set
Top 0.6%
1.5%
19
JMIR Medical Informatics
◐ 18 papers in training set
Top 0.6%
1.1%
20
Computer Methods and Programs in Biomedicine
○ 28 papers in training set
Top 0.7%
1.1%
21
IEEE Access
● 35 papers in training set
Top 1.0%
1.1%
22
Nucleic Acids Research
◐ 1281 papers in training set
Top 11%
1.1%
23
NAR Genomics and Bioinformatics
◐ 242 papers in training set
Top 4%
1.0%
24
Neuroinformatics
○ 46 papers in training set
Top 0.8%
0.8%
25
PLOS Digital Health
● 106 papers in training set
Top 4%
0.8%
26
Frontiers in Physiology
○ 106 papers in training set
Top 3%
0.6%
27
IEEE/ACM Transactions on Computational Biology and Bioinformatics
● 38 papers in training set
Top 1%
0.6%
28
IEEE Transactions on Computational Biology and Bioinformatics
● 20 papers in training set
Top 0.7%
0.6%
29
Journal of Computational Biology
○ 48 papers in training set
Top 1%
0.6%
30
Frontiers in Artificial Intelligence
○ 20 papers in training set
Top 0.9%
0.6%

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.