Summary of Taxonomy Tree Generation From Citation Graph, by Yuntong Hu et al.
Taxonomy Tree Generation from Citation Graph
by Yuntong Hu, Zhuofeng Li, Zheng Zhang, Chen Ling, Raasikh Kanjiani, Boxin Zhao, Liang Zhao
First submitted to arxiv on: 2 Oct 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 A novel end-to-end framework called Hierarchical Graph Taxonomy Learning (HiGTL) is proposed to enable automatic hierarchical taxonomy generation from citation graphs. This framework uses human-provided instructions or preferred topics as guidance and consists of a hierarchical citation graph clustering method that groups related papers based on textual content and citation structure, ensuring semantically meaningful and structurally coherent clusters. Additionally, the framework includes a novel taxonomy node verbalization strategy that iteratively generates central concepts for each cluster using a pre-trained large language model (LLM) to maintain semantic consistency across hierarchical levels. The HiGTL framework is designed with joint optimization to fine-tune both clustering and concept generation modules, aligning structural accuracy with the quality of generated taxonomies. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary HiGTL helps organize scientific knowledge by creating automatic hierarchies from citation graphs. This makes it easier to review literature and see emerging trends in research. The old way of doing this was slow, boring, and sometimes got things wrong. HiGTL uses a new method that groups papers based on what they say and how they’re connected. It also creates names for each group that make sense. This makes the results accurate and easy to understand. |
Keywords
» Artificial intelligence » Clustering » Large language model » Optimization
