Jing Liao

ASySD is a tool that efficiently removes duplicate citations in biomedical systematic reviews with high sensitivity and reliability.

This paper applies natural language processing to assess risk of bias in preclinical literature, achieving high F1 scores for several bias items using neural models and BERT.

The paper develops machine learning classifiers to automate citation screening in pre-clinical systematic reviews, achieving high sensitivity and specificity across different domains.

This paper applies machine learning to citation screening in systematic reviews of animal studies, improving sensitivity and identifying human errors.

This paper compares statistical methods for meta-analysis of animal studies, finding that normalized mean difference (NMD) has higher power than standardized mean difference (SMD), and meta-regression outperforms stratified meta-analysis.