Sentence BERT

Sentence BERT (SBERT) focuses on generating high-quality sentence embeddings, crucial for various natural language processing tasks like semantic similarity and text classification. Current research emphasizes improving SBERT efficiency through techniques like layer pruning and exploring its application in diverse domains, including depression detection, definition extraction, and even text-to-image synthesis, often employing ensemble methods or integrating SBERT with other architectures like CNNs or VAEs. These advancements enhance the accessibility and performance of SBERT, impacting fields ranging from mental health analysis to low-resource language processing and improving the accuracy and efficiency of various NLP applications.

Papers

March 29, 2022

Worldwide city transport typology prediction with sentence-BERT based supervised learning via Wikipedia
Srushti Rath, Joseph Y. J. Chow
LeArning Abstract Wikipedia Article Sentence BERT Urban Prediction City Typology Transportation Typology

February 8, 2022

Causal Scene BERT: Improving object detection by searching for challenging groups of data
Cinjon Resnick, Or Litany, Amlan Kar, Karsten Kreis, James Lucas, Kyunghyun Cho, Sanja Fidler
Autonomous Vehicle Raw Data Group Setting Perception Module Sentence BERT Perception Failure

January 12, 2022

PromptBERT: Improving BERT Sentence Embeddings with Prompts
Ting Jiang, Jian Jiao, Shaohan Huang, Zihan Zhang, Deqing Wang, Fuzhen Zhuang, Furu Wei, Haizhen Huang, Denvy Deng, Qi Zhang
Contrastive Learning Complex Prompt Sentence Embeddings Sentence Representation Prompt Embeddings Sentence BERT

November 30, 2021

Text Mining Drug/Chemical-Protein Interactions using an Ensemble of BERT and T5 Based Models
Virginia Adams, Hoo-Chang Shin, Carol Anderson, Bo Liu, Anas Abidin
Ticket BERT Relation Extraction Diverse Ensemble Text to Text Sentence BERT T5 Model Biomedical Text Mining Compound Protein Interaction

Sentence BERT

Papers

Worldwide city transport typology prediction with sentence-BERT based supervised learning via Wikipedia

Causal Scene BERT: Improving object detection by searching for challenging groups of data

PromptBERT: Improving BERT Sentence Embeddings with Prompts

Text Mining Drug/Chemical-Protein Interactions using an Ensemble of BERT and T5 Based Models