Summary of Empowering Interdisciplinary Research with Bert-based Models: An Approach Through Scibert-cnn with Topic Modeling, by Darya Likhareva et al.
Empowering Interdisciplinary Research with BERT-Based Models: An Approach Through SciBERT-CNN with Topic Modeling
by Darya Likhareva, Hamsini Sankaran, Sivakumar Thiyagarajan
First submitted to arxiv on: 16 Apr 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This novel approach leverages the SciBERT model and Convolutional Neural Networks (CNNs) to categorize academic abstracts from the Elsevier OA CC-BY corpus. The multi-segment input strategy processes abstracts, body text, titles, and keywords via BERT topic modeling through SciBERT. This method captures contextual representations using [CLS] token embeddings, concatenated and processed through a CNN. Class weights based on label frequency are incorporated to address class imbalance, significantly improving the classification F1 score. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Researchers have introduced a new approach to categorize academic abstracts from Elsevier’s OA CC-BY corpus. This method uses the SciBERT model and CNNs to stay current in their fields by processing abstracts, body text, titles, and keywords. The approach captures contextual representations using [CLS] token embeddings and class weights to address class imbalance. |
Keywords
» Artificial intelligence » Bert » Classification » Cnn » F1 score » Token
