Summary of Enriched Bert Embeddings For Scholarly Publication Classification, by Benjamin Wolff et al.
Enriched BERT Embeddings for Scholarly Publication Classification
by Benjamin Wolff, Eva Seidlmayer, Konrad U. Förstner
First submitted to arxiv on: 7 May 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Computation and Language (cs.CL)
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 The NSLP 2024 FoRC Shared Task I aims to develop a classifier that can predict one of 123 predefined research fields from the Open Research Knowledge Graph (ORKG) taxonomy for a given scholarly article. The paper presents its results, initially enriching the dataset and leveraging pre-trained language models like BERT, SciBERT, SciNCL, and SPECTER2 for transfer learning. The study explores feature-based and fine-tuned approaches, optimizing hyperparameters and investigating data augmentation from bibliographic databases. The results show that fine-tuning pre-trained models improves classification performance, with SPECTER2 being the most accurate model. Enriching the dataset with additional metadata also significantly improves classification outcomes. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper is about a competition to develop a computer program that can automatically classify research articles into different categories. This helps researchers find relevant resources more easily. The authors used special language models and combined them with extra information from other databases to make their program better. They tested many different combinations and found the best one, which was able to correctly classify articles most of the time. |
Keywords
» Artificial intelligence » Bert » Classification » Data augmentation » Fine tuning » Knowledge graph » Transfer learning
